<%BANNER%>

Record for a UF thesis. Title & abstract won't display until thesis is accessible after 2015-08-31.

DARK ITEM
MISSING IMAGE

Material Information

Title:
Record for a UF thesis. Title & abstract won't display until thesis is accessible after 2015-08-31.
Physical Description:
Book
Language:
english
Creator:
Letizia, Maria Cristina
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Degree:
Master's ( M.S.)
Degree Grantor:
University of Florida
Degree Disciplines:
Biomedical Engineering
Committee Chair:
Jiang, Huabei
Committee Co-Chair:
Ormerod, Brandi K
Committee Members:
Gunduz, Aysegul

Subjects

Subjects / Keywords:
Biomedical Engineering -- Dissertations, Academic -- UF
Genre:
Biomedical Engineering thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Statement of Responsibility:
by Maria Cristina Letizia.
Thesis:
Thesis (M.S.)--University of Florida, 2013.
Local:
Adviser: Jiang, Huabei.
Local:
Co-adviser: Ormerod, Brandi K.
Electronic Access:
INACCESSIBLE UNTIL 2015-08-31

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
lcc - LD1780 2013
System ID:
UFE0045990:00001

MISSING IMAGE

Material Information

Title:
Record for a UF thesis. Title & abstract won't display until thesis is accessible after 2015-08-31.
Physical Description:
Book
Language:
english
Creator:
Letizia, Maria Cristina
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Degree:
Master's ( M.S.)
Degree Grantor:
University of Florida
Degree Disciplines:
Biomedical Engineering
Committee Chair:
Jiang, Huabei
Committee Co-Chair:
Ormerod, Brandi K
Committee Members:
Gunduz, Aysegul

Subjects

Subjects / Keywords:
Biomedical Engineering -- Dissertations, Academic -- UF
Genre:
Biomedical Engineering thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Statement of Responsibility:
by Maria Cristina Letizia.
Thesis:
Thesis (M.S.)--University of Florida, 2013.
Local:
Adviser: Jiang, Huabei.
Local:
Co-adviser: Ormerod, Brandi K.
Electronic Access:
INACCESSIBLE UNTIL 2015-08-31

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
lcc - LD1780 2013
System ID:
UFE0045990:00001


This item has the following downloads:


Full Text

PAGE 1

INVIVOPHOTOACOUSTICMICROSCOPYOFSUBCUTANEOUSVASCULATUREINTHEHUMANWRIST:METHODBASEDONEXTERNALPRESSUREAPPLIEDTOTHEINVESTIGATEDREGIONByMARIACRISTINALETIZIAATHESISPRESENTEDTOTHEGRADUATESCHOOLOFTHEUNIVERSITYOFFLORIDAINPARTIALFULFILLMENTOFTHEREQUIREMENTSFORTHEDEGREEOFMASTEROFSCIENCEUNIVERSITYOFFLORIDA2013

PAGE 2

2013MariaCristinaLetizia 2

PAGE 3

Tomyfamilyandallwhohavebeensupportivetomeinmylife 3

PAGE 4

ACKNOWLEDGMENTS IwouldliketoacknowledgetheJ.CraytonPruittFamilyDepartmentofBiomedicalEngineeringattheUniversityofFloridaandthedepartmentofBiomedicalEngineeringatPolitecnicodiMilanoforgivingmetheopportunitytotakepartintheformativeAtlantisprogram.InparticularIacknowledgeProfessorHansVanOostrom,ProfessorGiuseppeBaselliandeveryonewhotookpartandmadepossiblethecreationoftheAtlantisCRISPdoubledegreeprogram.IwouldliketoexpressmygratitudetoProfessorHuabeiJiangforgivingmetheopportunitytoworkintheinterestingresearchareaofthebiomedicalopticalimaging.Thankstohistrust,inspirationandprecioussupportoverthisyear,Ihadtheopportunitytomakeprogressinmyresearchwork.MyappreciationgoestothecommitteemembersProfessorBrandiOrmerod,ProfessorAysegulGunduzforhavingacceptedtoexaminethisworkandforalltheirusefulremarkswhichcontributedtoimprovethevalueofthisthesis.IthankmylabmatesLeiXiandJianboTangfortheirhelpandforbeingsopatient.ItisthankstothemthatIhavelearnedmostofthetechniquesIhaveappliedinmyresearch.Awayfromthesit-downjob,aspecialthankstoValentinaandAlessandra,whosupportedmefromfaraway.Nomatterhowfarweare,friendshipovercomesanydistance.AveryspecialhugtoLuca,foralwaysbeingwithmeandfortheunforgettabletimespenttogetherinItalyaswellasintheUS.Thisthesiswouldnothavebeenpossiblewithouthisencouragement,positivespiritandlove.Finally,themostgratefulthankstomyparentsPaolaeNinoandtononnaMargheritafortheirloveandsupportthroughoutmylife.AstronghugtomysupercoolbrotherMarco,forhispatienceandforthetimehededicatedtome.Despitethedistance,youallhavealwaysbeenwithme. 4

PAGE 5

TABLEOFCONTENTS page ACKNOWLEDGMENTS .................................. 4 LISTOFFIGURES ..................................... 7 LISTOFABBREVIATIONS ................................ 11 ABSTRACT ......................................... 12 CHAPTER 1INTRODUCTIONANDTHEORY .......................... 14 1.1OverviewofPhotoacousticMicroscopy .................... 14 1.1.1PhotoacousticMicroscopyImagingFeatures ............. 17 1.1.1.1Photoacousticimagecontrast ............... 17 1.1.1.2Photoacousticimagepenetrationdepth .......... 19 1.1.1.3Photoacousticimagespatialresolution .......... 21 1.1.1.4Amplitudeofthepressurewaves .............. 22 1.1.1.5Imageformation ....................... 22 1.2PhotoacousticMicroscopyImagingofTumorVasculatureDevelopment:Motivations ................................... 22 1.2.1TumorAngiogenesis .......................... 23 1.2.2TheImportanceofImagingtheTumorMicrovasculatureDevelopment 24 1.2.3PhotoacousticMicroscopyImagingoftheVasculatureStructure .. 25 1.3PAMtoImagetheStructureoftheSubcutaneousVasculatureandAssessChangesoftheSignalGeneratedbyanExternalPressure ........ 27 2MATERIALSANDMETHODS ............................ 30 2.1Introduction ................................... 30 2.2SetUp,DataAcquisitionandImageProcessing ............... 30 2.2.1PAMSetUpandPAMeasurements .................. 31 2.2.2PostProcessingoftheMeasuredPASignal ............. 33 2.3PATheoryValidationwithPhantomExperiments .............. 34 2.3.1PhantomPreparation .......................... 35 2.3.2PhantomExperiments ......................... 36 2.4DesignofthePressureDevice ........................ 36 2.4.1MotivationsoftheDesignTopology .................. 36 2.4.2TechnicalDrawing ........................... 38 2.4.3SetUpDevelopedforthePressureApplication ........... 41 2.5InVivoPAMExperiments:ImagingoftheSubcutaneousVasculature ... 43 2.6RemovaloftheSignalFromtheSkin ..................... 47 2.6.1AutomaticAlgorithmforSkinProleDetectionandRemoval .... 49 2.6.2ImageSegmentationandRemovaloftheSkinSignal ........ 52 5

PAGE 6

3RESULTS ....................................... 56 3.1Introduction ................................... 56 3.2PATheoryValidationwithPhantomExperiments .............. 56 3.3InVivoPAMExperiments:SubcutaneousVasculature ........... 59 3.3.1FirstSetofExperiments ........................ 60 3.3.2SecondSetofExperiments ...................... 65 3.3.3ThirdSetofExperiments ........................ 69 3.3.4FourthSetofExperiments ....................... 73 4DISCUSSION ..................................... 79 5CONCLUSIONANDFUTUREDEVELOPMENTS ................ 88 REFERENCES ....................................... 91 BIOGRAPHICALSKETCH ................................ 96 6

PAGE 7

LISTOFFIGURES Figure page 1-1Principleofphotoacousticeffect. .......................... 17 1-2Absorptioncoefcientspectraofoxyhaemoglobin(redline),deoxyhaemoglobin(blueline),water(blackline),lipid(brownline),lipid(pinkline),melanin(blackdashedline),collagen(greenline),elastin(yellowline). ............. 20 1-3Absorptioncoefcientspectraofoxyhaemoglobin(redline),deoxyhaemoglobin(blueline). ....................................... 26 2-1SchematicofthePAMsystem. ........................... 31 2-2ImageformationinPAM. ............................... 32 2-3Flowchartoftheexperimentalprocedure,fromthemeasurementofthePAsignaltothereconstructionoftheimagedvolume. ................ 34 2-4Tumormimicphantom.Thetargetisthedarkspotlocatedinthecenterofthephantom.Thetargetissurroundedbytheagarosebackground. ........ 35 2-5Oneofthecalibrationweightsusedtoapplyaloadonthedevice,exertingapressureonthetarget. ................................ 38 2-6Squaredring-shapedpartofthepressuredevicewhichexertsthepressuredirectyonthetarget. ................................. 39 2-7Frontviewofthemodelofthebottompart. .................... 39 2-8Sideviewofthemodelofthebottompart. .................... 40 2-9Frontviewofthemodelofthetoppart. ...................... 40 2-10Modelofthewholestructure:thetoppartandthebottompartarexedtogetherandconnectedbythefourverticalcylinders. ................... 41 2-11Pictureofthewholestructure:thetoppartandthebottompartaremadeofasemi-transparentresinwithayellowishtintandarexedtogetherbythefourverticalcylindersmadeofsteel. ........................... 42 2-12Set-uparrangedtoexertthepressureonthetarget. ............... 44 2-13Positionofthepatientandofthepressuredeviceduringtherstexperiment.Anypressurewasexertedonthetarget. ...................... 46 2-14Positionofthepatientandofthepressuredevicewhenthepressureisexertedonthetargetregion.Theweightsareplacedintheplasticcontainerxedonthetopofthepressuredevice. ........................... 47 7

PAGE 8

2-15MIPalongthex)]TJ /F2 11.955 Tf 12.26 0 Td[(zplan.Thetoplayerrepresentstheskin,whilethebottomlayerrepresentsthesubcutaneousbloodvessels. ................ 48 2-16MIPalongthex)]TJ /F2 11.955 Tf 13.05 0 Td[(yplan.Theskinsignalisprojectedontothenalimage,makingthevasculaturebarelyvisible. ....................... 49 2-17Relationshipbetweenamplitudesofthesignalfromskinandbloodvessels,AsandAvrespectively. ................................ 50 2-18MIPalongthex)]TJ /F2 11.955 Tf 9.3 0 Td[(yplanaftertheskinremovalbyusingtheautomaticalgorithmimplemented. .................................... 52 2-19Sectionoftheimagedvolume.Theskinproleisshowninthewhitebottomlayer.Theotherwhitespotsovertheskinrepresentpartsofbloodvessels.Someseedpoints,red-colored,havebeenplacedalongtheskinprole. ... 53 2-20Sectionoftheimagedvolume.Theseedpointsgrewandtooktheshapeoftheskin. ........................................ 54 2-21MIPalongthex)]TJ /F2 11.955 Tf 11.96 0 Td[(yplanafterthesegmentationandtheskinremoval. .... 55 3-1Tumormimicphantomexperimentalresultwhentheabsorptioncoefcientofthetargetis0.07mm)]TJ /F6 7.97 Tf 6.59 0 Td[(1. ............................... 58 3-2Tumormimicphantomexperimentalresultwhentheabsorptioncoefcientofthetargetis0.049mm)]TJ /F6 7.97 Tf 6.58 0 Td[(1. .............................. 58 3-3Tumormimicphantomexperimentalresultwhentheabsorptioncoefcientofthetargetis0.021mm)]TJ /F6 7.97 Tf 6.58 0 Td[(1. .............................. 59 3-4MIPofthestructureofthebloodvesselsintherstsetofexperiments.Thelocatedbloodvesselsareenumeratedfrom1to7.Lettersareusedtoindicatethebranchesofabloodvessel. ........................... 60 3-5MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto0N. .... 63 3-6MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto7000N. .. 63 3-7MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto15000N. 64 3-8MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto20000N. 64 3-9MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto25000N. 65 8

PAGE 9

3-10MIPofthestructureofthebloodvesselsinthesecondsetofexperiments.Thelocatedbloodvesselsareenumeratedfrom1to6.Lettersareusedtoindicatethebranchesofabloodvessel. ...................... 66 3-11MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto0N. .... 67 3-12MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto7000N. .. 68 3-13MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto15000N. 68 3-14MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto20000N. 69 3-15MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto25000N. 69 3-16MIPofthestructureofthebloodvesselsinthethirdsetofexperiments.Thelocatedbloodvesselsareenumeratedfrom1to7. ................ 70 3-17MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto0N. .... 71 3-18MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto7000N. .. 72 3-19MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto15000N. 72 3-20MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto20000N. 73 3-21MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto25000N. 73 3-22MIPofthestructureofthebloodvesselsinthefourthsetofexperiments.Thelocatedbloodvesselsareenumeratedfrom1to10. ............... 74 3-23MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto0N. .... 76 3-24MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto7000N. .. 76 3-25MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto15000N. 77 9

PAGE 10

3-26MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto20000N. 77 3-27MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto25000N. 78 10

PAGE 11

LISTOFABBREVIATIONS ARPAMAcoustic-ResolutionPhotoacousticMicroscopyHbHaemoglobinHbO2DeoxyhaemoglobinMIPMaximumIntensityProjectionNIRNearInfraredNONitricOxidePAPhotoacousticPAMPhotoacousticMicroscopyROIRegionofInterestUSUltrasoundVEGFVascularEndothelialGrowthFactor 11

PAGE 12

AbstractofThesisPresentedtotheGraduateSchooloftheUniversityofFloridainPartialFulllmentoftheRequirementsfortheDegreeofMasterofScienceINVIVOPHOTOACOUSTICMICROSCOPYOFSUBCUTANEOUSVASCULATUREINTHEHUMANWRIST:METHODBASEDONEXTERNALPRESSUREAPPLIEDTOTHEINVESTIGATEDREGIONByMariaCristinaLetiziaAugust2013Chair:HuabeiJiangMajor:BiomedicalEngineeringTheimagingofthevasculatureprovidesapowerfultoolforstudyingthevasculaturedevelopmentaswellasassessingstructuralandfunctionalfeaturesofbloodvessels.Thiskindofimagingbecomesmuchmoreimportantifitisaddressedtothestudyoftumorangiogenesis.Tumorangiogenesisistheproliferationofanetworkofbloodvesselsthatpenetratesintocancerousgrowths,supplyingnutrientsandoxygenandremovingwasteproducts.Adequatenon-invasiveimagingtechniquescanhelpphysicianstomonitorandcontrolthetumorvasculaturedevelopmentaswellastoassessfunctionalandstructuralcharacteristicsofthetumorvasculature.Apotentialnon-invasiveimagingmodalityisphotoascoustic(PA)imaging,anewbiomedicalimagingmodality,thatisemergedoverthelastdecade.Itconsistsinimagingoftheinternaldistributionofopticalenergydepositioninbiologicaltissuesbasedonthedetectionoflaser-inducedultrasonicwaves,whichrevealsphysiologicallyspecicopticalabsorptioncontrast.IndeedaPAimageisanultrasoundimageinwhichthecontrastdependsnotonthemechanicalandelasticpropertiesofthetissue,butbyitsopticalproperties,particularlyopticalabsorption.Thus,thisallowsPAimagingtovisualizeanatomicalfeaturesthatcontainanabundanceofchromophoressuchashaemoglobin,deoxyhaemoglobin,melanin,lipidsandwater,whichabsorblight. 12

PAGE 13

Inthiswork,photoacousticmicroscopy(PAM)hasbeenusedtoimagethestructureofvasculatureandalsotoevaluatehowthiskindofacquisitionismodiedbytheapplicationofanexternalpressureonthetargetregion.Apressuredevicehasbeendesignedandmanufacturedtoapplyanexternalmechanicalpressuretotheinvestigatedregion.Firstly,weimagedthestructureofsubcutaneousvasculatureinthehumanwristwithPAMinabsenceofpressure.ThenweperformedPAMexperimentsonthesameareawhenanexternalpressurewasapplied.Inbothcases,wechose532nmlaserlight,whichisanisosbesticwavelengthforhaemoglobinanddeoxyhaemoglobin.Finally,theeffectsgeneratedbytheexternalpressurewereassessedintermsofchangesofowrateandbloodvolume,bymeasuringchangesintheamplitudeofthePAsignal.Theresultsofthisworkshowthattheresponseofthebloodvesselstothepressureapplicationcanbedividedintotwogroups.Forcertainvessels,thebloodvolumeandbloodowprogressivelydecreasewhenthepressureisexerted;thepressureapplicationcausesanincreaseinthebloodowandvolumeinothervessels.Otherin-vivoexperimentsshowthatthebloodvolumeincreasesforcertainvaluesofexternalpressureanddecreaseforothers.Theincreaseoftheamplitudeofthesignalimpliesanenhancementofthenalimage,sincetheshapeofthebloodvesselsisshownwithabetterdenitionandthesignalfromthevesselsbecomesstrongerthanthatfromthesurroundingtissue.Althoughtheconceptofusinganexternalpressuretoimprovethequalityoftheimagingwasdemonstratedinthisthesis,fewotheraspectsconcerningtheorganizationofourexperimentsmightbeinvestigatedinfuturestudies.Infutureexperiments,beforeapplyingahigherexternalpressureandstartanewacquisition,enoughtimemightbelefttothecirculationtoreturntoitsrestcondition.Acomparisonofthecurrentresultsandthenewresultsmightbeinteresting. 13

PAGE 14

CHAPTER1INTRODUCTIONANDTHEORYInthischapterthegoalandthepurposeofthisworkaredescribed.Ononehand,anoverviewabouttheevolutionofthephotoacousticmiscroscopytechniqueisprovided,fromthediscoveryofthephotoacousticeffectbyBellinthe1880stotheresearchactivitiesbasedonthisphenomenon.Afterwards,thephysicalprinciplesbehindthephotoacousticeffectareexplained,toshowhowultrasoundwavescanbegeneratedbylaserlight.Then,thefeaturesofphotoacousticmicroscopyimagingareprovided,bycomparingsuchtechniquetotraditionalultrasoundimagingandpurelyopticalimaging.Thistheorywillbeextensivelydeepenedinsections 1.1.1 ,wheretheadvantageousfeaturesofPAMimagingaredescribedtounderstandthiswork.Ontheotherhand,theeldofstudywhereourworkisplacedisillustrated:abriefdescriptionofthetumorangiogenesisisprovidedinsection 1.2.1 .Theimportanceoftheimagingofthetumorvasculatureisintroducedanddeepenedinsection 1.2.2 .Afterwards,weexplainhowphotoacousticmicroscopymeetstheneedforsuchkindofimaging.Particularly,weshowthewaytheimagingfeaturespreviouslydescribedcanbeexploitedtosuccesfullyimagethestructureofthevasculature.Finally,weillustratesomerelevantPAMresultstoimagethestructureofthevasculature,frompreviousstudiesinourlaboratoryandtheliterature.Then,wefocusonthedescriptionofmotivationsandpurposesofourstudy.TheinnovationofthisworkconsistsinimagingthestructureofthesubcutaneousvasculaturewithPAMaswellasinevaluatingtheeffectsonthePAMimagewhenanexternalpressureisappliedonthetargetregion. 1.1OverviewofPhotoacousticMicroscopyThehistoryofthephotoacoustictechniquesdatesbackto1880,whenAlexanderGrahamBelldiscoveredthephotoacousticeffect.Thiseffectoccurswhenaperiodicallyinterruptedbeamofsunlightshinesonasolidinanenclosedcell:anaudiblesoundcouldbeheardbymeansofahearingtubeattachedtothecell.Theabsorbedenergy 14

PAGE 15

fromthelightcauseslocalheatingandthroughthermalexpansionapressurewavetoo[ 1 ].ThetwophysicistsJohnTyndallandWilhenmRoentgen,encouragedbyBell'sdiscoveries,foundthatanacousticsignalcanalsobegeneratedwhenagasinanenclosedcellisilluminatedwithchoppedlight[ 2 3 ].Bellsubsequentlyperformedexperimentswithavarietyofsolids,liquids,andgasesbutthepracticaldevelopmentswerepoorandphotoacoustictechniquesweretemporarilyignored.Fiftyyearslater,thephotoacousticeffectwithgaseswasagaintakenintoaccountalthoughthePAeffectwithsolidshasneverbeenvalidated[ 4 ].Bellattributedthephotoacousticeffectobservedwithspongysolidstoacyclicreleaseofairpulsesfromtheporesofthesolid,duetothecyclicalheatingandcoolingofthesolidbythechoppedlight[ 4 5 ].HealsosupportedthetheoryofRayleighwhorelatedthephotoacousticeffecttothemechanicalmotionofthesolid.However,accordingtoPreeceexperiments,thesoliddoesnotundergoanysubstantialmechanicalmotionandtheeffectisduetoanexpansionandcontractionoftheairinthecell[ 6 ].Followingthisline,duringthe1880sMercadierconcludedthatthesoundisduetoVibratorymovementdeterminedbythealternateheatingandcoolingproducedbytheintermittentradiations,principallyinthegaseouslayeradheringtothesolidsurfacehitbytheseradiations,completingthestudiescarriedonbythepreviousscientists[ 7 ].Thereafter,theresearchactivitiesweremoderatedbothattheuniversityandindustriallevel.Withthedevelopmentofthelaser,in1960s,anhighpeakofpowercouldbeconnedinanarowbeamwithaverypurespectralpurity,satisfyingtheneedsofmanyPAsensingapplications[ 8 ].IncludingthelasertechnologyinthePAM,industrialapplicationofthephotoacousticeffectapparedinthe1970sand1980s.Theseapplicationsgenerallyexploitedtheindirectgas-phasecelltypeofPAdetection,inwhichacousticwavespropagatinginagasgeneratedbylaser-inducedsurfaceheatingaredetectedwithamicrophone.Thisisincontrasttothedirectdetectionoflaser-inducedultrasoundwaveswhichbiomedicalPAimagingexploits.Thisdirectdetectionapproach 15

PAGE 16

wasinitiallyexploredforcharacterizingsolidsasapotentialnon-destructivetestingtool.Onlyduringthemid-1990s,itbegantobeinvestigatedforbiomedicalimagingalthoughtherstimagesbegantoappearthereafter.Thisearlywork,undertakenbyahandfulofresearchers,progressedsteadilyuntiltheearlytomid-2000swhenthersttrulycompellinginvivoimagesreachedtheirmaturity.Fromthispointonwards,theeldhaswitnessedmajorgrowthintermsofthedevelopmentofinstrumentation,imagereconstructionalgorithms,functionalandmolecularimagingcapabilitiesandtheinvivoapplicationofthetechniqueinclinicalmedicineandbasicbiologicalresearch.Inphotoacousticimaging,ultrasoundwavesareexcitedbyelectromagneticradiation.Duringphotoacousticexperiments,shortpulsedlaseriscommonlyusedtogenerateultrasoundwavesatopticalwavelenghtsinthevisibleandnear-infrared(NIR)partofthespectrum,whereNIRspectrumrangeliesinopticaltransparentwindowprovidingthegreatestpenetrationdepth.BecauseoftheuseofvisibleorNIRlight,thereisnoradiationissuecomparedwithconventionalX-rayimagingtechniques[ 9 ].Forphotoacousticmicroscopy,thePAimagingisobtainedthroughmechanicallyscanningofafocusedtransducer.Sinceafocusedtransducerwithoutlightbeamfocusingisemployed,itistermedacoustic-resolutionmicroscopy(ARPAM)[ 8 ].AsshowninFig. 1-1 ,absorptionbyspecictissuechromophores,suchasHb,melanin,H2O,lipidsproducesasmallthermalexpansion(lessthan0.1K),withoutdamagingtissues.Thisleadstoaninitialpressureincrease.Thesubsequentrelaxeationresultsintheemissionofbroadband(tensofMHz)lowamplitude(lessthan10kPa)acousticwaves[ 10 ].ThesewavespropagatetothesurfacewherethePAsignalsaredetectedbyhighly-sensitiveultrasoundreceiversinordertoacquireasequenceofA-lines.Bymeasuringthetimeofarrivaloftheacousticwavesandknowingthespeedofsound,thedistribituionofopticalabsorptionsinthesample,inotherwordstheimage,canbereconstructed[ 11 ].ARPAMworksinreection(orbackward)mode,ratherthan 16

PAGE 17

Figure1-1. Principleofphotoacousticeffect. orthogonalortrasmissionmode.Itprovidesreal-timeA-scansandtheacquiredA-linesareusedtoformanimagewithoutanyreconstructionalgorithms[ 8 ]. 1.1.1PhotoacousticMicroscopyImagingFeaturesThedescriptionoftheimagingfeaturesofPAMisimportanttofullyunderstandtheabilitiesofthistechniqueandistreatedinthissection,whereasthewaythisfeatureswillbeexploitedinthepresentworkwillbeexplainedinsection 1.2.3 .ForthedescriptionofthemostadvantageousfeaturesofPAimaging,acomparisonbetweenultrasound(US)imagingandPAimagingisalsoprovided. 1.1.1.1PhotoacousticimagecontrastThesignalgenerationinPAMisakeyelementinunderstandingthesourceofcontrast.Thepulsedlaserlightilluminatesthetissuesurfaceandisabsorbedby 17

PAGE 18

moleculesknownaschromophores.Theabsorbedenergyprovokesvibrationalaswellascollisionalrelaxation,andisconvertedtoheatafterward.Thisproducesaninitialpressureincrease[ 10 ].Theinitialpressuredistributionpoisencodedontopropagatingacousticwaveswhicharedetectedbyanultrasoundtransducerlocatedonthesurfaceandthenconvertedtoatime-resolvedelectricalsignal.ThePAimageisthenformedfromasetofthesePAsignalsdetectedatdifferentpoints,thusitisarepresentationoftheinitialpressuredistributionpo.poisrelatedtotheheatproducedbythelaserenergy.Fromsimplethermodynamicconsiderations[ 8 ],thepressurepoislinkedwiththeabsorbedopticalenergyH(r)byEqn. 1 :po(r)=)]TJ /F3 11.955 Tf 30.25 0 Td[(H(r) (1)where)]TJ /F1 11.955 Tf 10.1 0 Td[(istheGruneisenconstant,whichrepresentsameasureoftheconversionefciencyofheattopressureenergyanditisdenedas)-371(=c2=Cp,whereisthevolumethermalexpansivity,cisthesoundspeedandCpthespecicheatcapacityataconstantpressure.TheabsorbedopticalenergyH(r)isgivenbytheproduct:H(r)=a(r)(r;a,s,g) (1)whereaandsaretheabsorptionandscatteringcoefcientsrespectively,gistheanisotropyfactorandistheopticaluence.TheEqn.( 1 )canbethenobtainedbywritingtheEqn.( 1 )explicitly:po(r)=)]TJ /F3 11.955 Tf 30.26 0 Td[(a(r)(r;a,s,g). (1)Thustheinitialpressuredistributionpoisdeterminedbytwoparts:amechanicalandthermodynamiccontributionrepresentedbytheGruneisencoefcient)]TJ /F1 11.955 Tf 10.1 0 Td[(and 18

PAGE 19

anopticalcontribution.Sincethemechanicalandthermodynamicpropertiesareusuallythoughttovarysufcientlyweaklybetweendifferenttissuetypes,theycanbeconsideredspatiallyinvariant.Consequentlytheinitialpressuredistributionisdeterminatedbytheopticalabsorptionandscatteringpropertiesofthetissue.Furthermore,sincetheopticalabsorptionhasthemostimportantinuencetothecontrast,thePAimagingisoftendescribedasabsorption-based.Forthesakeofprecision,poisanonlinearfunctionofa,becauseitisproportionaltotheproductofaandwhichisitselfdependentona[ 8 ].ThefactthattheamplitudeofthePAsignalisproportionaltotheabsorbedopticalenergydistribution,impliesanimportantadvantagethatthePAimagingprovidescomparedtoUSimaging:theformerprovidesagreatertissuedifferentiationandspecicitythanthelatter,becausethedifferencesinopticalabsorptionbetweendifferenttissuesaremuchlargerthanthoseinacousticimpendance.Fromtheprevioustheoreticalanalysis,itemergesthatthecontrastsdependsontheabsorptionbymoleculescalledchromophores.Themostimportantchromophoresinthevasculaturearehaemoglobin,deoxyhaemoglobin,melanin,lipidsandwater.Theabsorptionpropertiesofeachchromophorearewavelength-dependent,asFig. 1-2 shows.Theknowledgeoftheabsorptionspectrumofeachchromophoreisakeyelementinthechoiceoftheexcitationwavelength.Thisapproachwillbeextensivelyadoptedinthisworkinsection 2.5 1.1.1.2PhotoacousticimagepenetrationdepthOpticalandacousticalattenuationsdenethepenetrationdepths.Insofttissues,theultrasonicpenetrationdepthisinverselyproportionaltotheultrasonicfrequency,becausetheacousticattenuationcoefcientisproportionaltotheultrasonicfrequency.Howeveritisultimatelytheopticalattenuationthatdominates[ 8 ].Suchattenuationdependsontheabsorptionandscatteringcoefcients,aandsrespectively,anditisverywavelenght-dependent,asFig. 1-2 shows.Accordingtodiffusiontheory,the 19

PAGE 20

Figure1-2. Absorptioncoefcientspectraofoxyhaemoglobin(redline),deoxyhaemoglobin(blueline),water(blackline),lipid(brownline),lipid(pinkline),melanin(blackdashedline),collagen(greenline),elastin(yellowline).Figureadaptedfrom[ 13 ]. penetrationdepthisbestcharacterizedbye,whichisdenedase=p 3a(a+0s) (1)where0sisthereducedscatteringcoefcient.Inhomogeneousscatteringmedia,whenthelightreachesadepthbeyondseveraltransportmeanfreepaths(about1mm),thedescriptionofphotonpropagationchangesfromtheballisticregimetothediffusiveregime[ 9 14 15 ].Consequently,theirradiancedacaysexponentiallywithdepth,withaspaceconstante.Thismeansthat1=eisthepenetrationdepth,atwhichtheirradianceis1=e[ 8 ].ThislimitatedpenetrationdepthrepresentsoneofthemostchallengingfeatureofthePAimaging,becauseasthelightpenetratesseveralcentimetresintissue,itdeterminesasignalattenuationofseveralordersofmagnitude.Nevertheless,ithasbeenshownthatstructuresatadepthbeyondseveralcentimetreshavebeenimaged,byproperlyoptimizingthelightdelivery,thetransducerdesignparametersandthesignalprocessingandwiththeuseofcontrastagentstoreducetheweaknessoftheultrasoundsignal[ 10 ]. 20

PAGE 21

IfwecomparethistechniquewithUSandpurelyopticalimagingtechniques,thepenetrationdepththatPAimagingcanachieve,isoneorderofmagnitudelowerthantheoneachievedbyUSimaging,whichcanbe10cmorevenmoreinsofttissues[ 18 ].However,thepenetrationdepthavailabletoPAimagingsignicantlyexceedsthatofpurelyopticalimagingtechniquessuchasmultiphoton,confocalmicroscopyoropticalcoherencetomographythatrelyonunscatteredorso-calledballisticphotons[ 19 20 ].Inthesetechniques,thediffusionphenomenalimitthepenetrationto1mminthehumantissue,duetoitshigh-scatteringproperties.Onthecontrary,inPAimagingthephotonsdiffuseinsidethescatteringtissuetoo,eventhough,allthephotonsarrivingatthetargetsareusefulphotonswhichwillbeabsorbedbythetargets.Thenmostoftheabsorbedenergyisconvertedtoacousticsignals[ 21 ].Theunderstandingofthewavelength-dependentnatureofthepenetrationdepthwillbeusedinthisworkinchapter 2 tochoosethewavelengthoftheexcitationlighttooptimizethelightdeliveryandtoestimatethespatialresolutionaccordingtothedepthofthetarget. 1.1.1.3PhotoacousticimagespatialresolutionThespatialresolutionisdenedbytheultrasonicparametersandtheimagingdepth.Thelateralresolutionisdeterminedbythecenterfrequencyofthetransducerandtheapertureofthetransducer.Itsvalueisgivenby0.61(0=NA),where0isthecenteracousticwavelengthandNAisthetransducernumericalaperture,i.e.thesolidanglesubtendedbythetransducerapertureandthepointontheimagedvolume.Theaxialresolutionisprimarilydeterminedbythefrequencybandwidthofthetransducerdetector.Therefore,toachievehighspatialresolution,atransducerwithlargenumericalaperture,highcenterfrequencyandawidebandisnecessary[ 9 21 ].Thecenterfrequency,however,islimitedbythetargetimagingdepth.Infact,themaximumfrequencycontentofthePAwavesisusuallynotlimitedbythegenerationprocessitself,butbythebandlimitingofthepropagatingPAduetotheacousticattenuation[ 8 22 ]. 21

PAGE 22

Theacousticattenuationstronglydependsonthetissuetype;however,ingeneralwithapenetrationdepthoffewhundredmicrometers,asub-10mspatialresolutionisachievable.Ifthepenetrationdepthincreasestomillimetres,theattainablespatialresolutionislessthan10manditbecomeslowerthan1mmforcentimetrepenetrationdepth[ 9 21 ].Byconsequence,thespatialresolutionscaleswithdepth.Theknowledgeofthephysicalprinciplewhichdeterminethespatialresolutionisakeyelementduringtheprojectofourexperiments.Thistheorywillbeappliedinsection 2.3.1 andinsection 2.5 ,duringthechoiceofthepropertransducer,toachievethedesiredspatialresolutionandlateralresolutionaccordingtothedimensionofthetarget. 1.1.1.4AmplitudeofthepressurewavesTheamplitudeofthepressurewavesgeneratedbyphotoacousticimagingislessthan10KPa,whereasthepressurewavesinvolvedinPAexperimentsexceed1MPa[ 18 ]severalordersofmagnitudehigherthenUSwaves.Thisimpliesasatefy-relatedconsiderationsincethepotentialhazardsduetoUSexposureareabsentinPAexperiments. 1.1.1.5ImageformationAnotherdifferencebetweenUSandPAimagingreliesontheimageformation.Intheformer,localizationcanbeachievedbyfocusingboththetransmitbeamandthereceivebeam[ 18 ].Inthelatter,fordepthsgreaterthanapproximately1mm,onlythereceivebeamcanbefocused,becausetheopticalscatteringexhibitedbymostsofttissuespreventsthetransmitbeam,i.e.theexcitationlight,tobefocusedandthenlocalized. 1.2PhotoacousticMicroscopyImagingofTumorVasculatureDevelopment:MotivationsInthissectionweillustratethescenariowherethepresentworkisplaced.Abriefdescriptionofthetumorangiogenesisisrstlyprovidedinthesection 1.2.1 wherethemostimportantmechanismsbehingthisphenomenonareexplained.Then,thistheory 22

PAGE 23

willbeusedinsection 1.2.2 toexplaintheimportantneedforanon-invasiveimagingofthetumorvasculature.Insection 1.2.3 ,theimagingabilitiesdescribedinsection 1.1.1 willbeusedtoillustratehowthestructureofthevasculaturecanbesuccesfullyimagedbyPAM. 1.2.1TumorAngiogenesisTumorangiogenesisistheproliferationofanetworkofbloodvesselsthatpenetratesintocancerousgrowths,supplyingnutrientsandoxygenandremovingwasteproducts[ 36 ].Thisphenomenonoccourswhentumorgrowstoasizeofover1-2mm3,whenhypoxiaandnutrientdeprivationbecomeimportantandthetumorneedstoevacuatemetabolicwasteaswellascarbondioxide[ 36 37 ].Toallowthetumortoprogress,theangiogenicswitchisactivated.Thus,thetumorattractsbloodvesselsfromthesurroundingstroma.Endothelialcellsdetachtheirjunctionaladhesionfromtheirneighbors,sprouttowardsgradientsofproangiogenicfactors,proliferatetoformprovisionaltubes,recruitperivascularcellstoprovidestabilityandremodeltoformafunctionalnetwork[ 37 38 ].Tumorangiogenesisisinvolvednotonlyinprimarytumors,butalsoinmetastasisformationandfurtheroutgrowthofmetastases.Thereisawiderangeofcelltypesandsignalsinvolvedinangiogenesis.Theunderstandingthewholeangiogenicprocessiscurrentlyaddressingresearchtowardstheuseofanti-angiogenicdrugstotargetandinhibitspecicfactorsinvolvedintheremodeling.Membersofvascularendothelialgrowthfactor(VEGF)familyarewell-knownangiogenesisactivators[ 31 42 43 ].Despitethepromisingactivityofanti-angiogenicdrugsinpreclinicaltumormodels,targetingVEGFsignalingappearstobeinsufcientforpermanentlyinhibitingtumorangiogenesisinpatientswithcancers.Infactithasbeenseenthatmanytumorscanovercometheuseofangiogenicinhibitors,byactivatingorupregulatingalternativeproangiogenicpathwaysaftertherstpathwayisinhibited[ 36 39 ].Thus,identifyingwhichstepsandcomponentsofthisprocessarethemostsusceptibletotreatmentwithdrugsisanimportantconcern.Themosteffectivetreatmentswillprobablyinvolve 23

PAGE 24

targetingcombinationoffactors,aswellasimprovingtheefciencyofdrugdeliverytothetumormicroenvironment. 1.2.2TheImportanceofImagingtheTumorMicrovasculatureDevelopmentAlthoughangiogenesismightinitiallyprovidethetumorwithmoreoxygenandnutrients,ultimatelytheresponseispoor.Infact,thecontinuoslyremodeledtumorvasculatureisleakyandtortuousandthenewvesselsdonotbecomematurevesselswithproperfunctions[ 36 38 41 ].Angiogenesiswithinmalignanttumorsisadisorganisedandchaoticprocess.Manydifferentfeaturesofvascularitypermitthedistinctionbetweenmalignantandbenignprocesses,someofwhichcanbeinterrogatedbyimagingtechniques[ 44 45 ].Structuralandfunctionalcharacteristicsofmalignanttumorvasculatureinclude: Spatialheterogeneityandchaoticstructure.Somebloodvesselsareexposedtopersistentangiogenicstimulatorysignalsmorethanotherswithinthetumormicroenvironment.Asaconsequence,thevasculardensityisextremeheterogeneous,withareasoflowvasculardensitymixedwithregionsofhighangiogenicactivity[ 37 ]. Poorlyformed,fragilevesselswithhighpermeabilitytomacromolecules.Vesselwallstructureisabnormalintumors.Infact,largeinterendothelialjunctions,increasednumbersoffenestrations,vesiclesandvesicovacuolarchannels,andalackofnormalbasementmembraneareoftenfoundintumorvessels.Inagreementwiththesestructuralalterationsinthetumorvesselwall,vascularpermeabilityofsolidtumorvesselsisgenerallyhigherthanthatofmostnormalvessels.Extravasationofmoleculesfromthebloodstreamoccursbydiffusion,convection,and,tosomeextent,bytranscytosisinanexchangevessel[ 36 ]. Arterio-venousshunting,highvasculartortuosityandvasodilatation.Thenormalmicrovesselsconsistofdifferentiatedunitssuchasarterioles,capillariesandvenules,andformawell-organizedarchitecturewithdichotomousbranchingandhierarchicorder.Incontrast,tumorvesselsaredilated,saccular,tortuous,andheterogeneousintheirspatialdistribution.Moreover,tumorvasculatureisdisorganizedandhastrifurcationsandbrancheswithunevendiameters[ 46 ]. Intermittentorunstablebloodow.Bloodperfusionintumorsisspatiallyandtemporallyheterogeneous.Arteriovenouspressuredifferenceandowresistancegovernbloodowinavascularnetwork.Flowresistanceisafunctionofgeometric(vasculararchitecture)andviscous(bloodviscosity,rheology)resistances. 24

PAGE 25

Abnormalitiesinbothvasculatureandviscosityincreasetheresistancetobloodowintumors[ 38 ].Focalleaks,whichoftenexistinsomeofthetumorvessels,mayalsocompromisethedownstreambloodow[ 41 ]. Abnormalmetabolicenvironmentintumor.Hypoxiaandacidosisarethehallmarksofadysfunctionalmetabolicenvironmentinsolidtumors.Owingtotheirabnormalstructureandfunction,tumorvesselsareunabletodeliveradequatelevelsofnutrientsandoxygentotumorsandtoremoveacidicwasteproductsoutoftumors[ 37 ].Localimbalanceofangiogenesisandtumorcellproliferationmakeshypovascularregionsintumors.Theintermittentbloodowcausesperiodicalhypoxiainatumorwhichiscalledacuteorperfusion-limitedhypoxia[ 38 ].Eventhepresenceofbloodowdoesnotguaranteethedeliveryofoxygeninsolidtumors[ 36 41 47 ].Severalcurrentimagingmodalitieshavedifferentlimitationsformonitoringvasculaturedevelopment.X-rayComputedTomography(CT)needsextrinsiccontrastagentandexposespatientstoionizationradiation[ 31 ].AlsoPositronEmissionTomography(PET)screeningisinvasiveasofteninvolvestheuseofextrinsiccontrastagents.Ontheotherhand,MagneticResonanceImaging(MRI)isnoninvasive,butislimitedbyitslowtemporal-spatialresolution.Purehigh-resolutionopticalimagingmodalitiessuchassingle-photon,multi-photonuorescencemicroscopysufferfromlimitedimagingdepth(lessthan1mm)andrepeateduorescentdyeinjection[ 31 ].Adequatenon-invasiveimagingtechniquescanhelpphysicianstomonitorandcontrolthetumorvasculaturedevelopment,toassessfunctionalandstructuralcharacteristicsofthetumorvasculaturetodistinguishbetweenabenignandmalignanttumor.Theseimagingtechniqueswouldalsobeusefultodeterminewhetherandwheretostartananti-angiogenicorchemo-orradiotheraphytreatment,tocontrolthetumorgrowthandalsotheresponseofthetumortotheanti-angiogenictreatment.Onepotentialnon-invasiveimagingmodalityisphotoascousticimaging(PA)[ 17 21 33 34 48 49 ]. 1.2.3PhotoacousticMicroscopyImagingoftheVasculatureStructureInthissection,wedescribewhyandhowastructuralimageofthevasculaturecanbeobtainedwithPAMimaging. 25

PAGE 26

Figure1-3. Absorptioncoefcientspectraofoxyhaemoglobin(redline),deoxyhaemoglobin(blueline).Figureadaptedfrom[ 51 ]. Asalreadyexplainedinsection 1.1.1 ,thecontrastdependsontheabsorptionpropertiesofthechromophoresinthevasculature.Eachchromophoreischaracterizedbywavelength-dependentabsorptionproperties.Thustheanalysisoftheabsorptionspectraofthechromophorehastobeperformed,beforechoosingtheopticalwavelengthsfortheimagingpurpose.Fig. 1-3 showstheabsorptioncoefcientspectraofthehaemoglobinanddeoxyhaemoglobin.Haemoglobinanddeoxyhaemoglobinaretreatedasthemajoropticalabsorbersintheblood.Toobtainastructuralimageofthevasculature,thetotalhaemoglobinconcentrationhastobeconsideredbychoosinganisosbesticwavelength,atwhichhaemoglobinanddeoxyhaemoglobinexhibitthesamestrongmolarabsorptioncoefcient[ 21 23 50 ].Bydoingso,theabsorptionisindependentofoxygenationandchangesinabsorptionresultfromchangesinthetotalhaemoglobinconcentrationorbloodvolume.Moreover,toachievehighimagecontrast,theopticalabsorptionofbloodhastobestrongcomparedwiththatofthesurroundingtissue[ 16 32 52 53 ]. 26

PAGE 27

Withthisapproach,thebloodvesselsspatialdistributioncanbeimaged,theshapeandstructureofthevasculaturecanbeobservedandthebloodvesselvolumecanbeassessed. 1.3PAMtoImagetheStructureoftheSubcutaneousVasculatureandAssessChangesoftheSignalGeneratedbyanExternalPressureInthissection,theimportanceofimagingthestructureofthevasculatureisbrieypointedout.Secondly,somerelevantPAMresultstoobtainstructuralimagesaredescribedfrompreviousstudiesinourlaboratoryandtheliterature.Finally,wefocusonthedescriptionofreasonsandpurposesofourexperiments.Whenimagingthemicrovasculature,especiallythegrowingtumormicrovasculature,animportantfeaturethathastobeassessedregardsthestructureandtheshapeofthebloodvessels.Asalreadydescribedinsection 1.2.2 ,thespatialdistributionoftumorbloodvesselsisheterogeneousandareasorhighvasculardensityaremixedtoareaswithlowangiogenicactivity.Moreover,thearchitectureofbloodvesselsisnotwell-organized.Branches,suchasarterioles,venulesorcapillaries,areoftennotdifferentiatedaccordingtothephysiologichierarchicorderandpresentunevendiameter.Anotherfundamentalfeaturethathastobeinvestigatedwhendealingwithmicrovasculature,andparticularlywithtumormicrovasculature,isthebloodowandespeciallythechangesinbloodow.Thebloodowaroundtumorsmaybeunstableandspatiallyandtemporallyheterogeneous,duetoaforementionedabnormalitiesintheshapeandstructureofthebloodvesselsandintherheologicpropertiesofthebood.Asmotivatedinsection 1.2 ,theimagingofthestructureofthevasculatureisnecessarytoobtainanatomicalandhaemodynamicalinformationaboutbloodvessels.Insection 1.2.3 weexplainthataninvestigationofthemorphologyofbloodvesselsbyusingPAMispossible.ThisPAMabilitywasexploitedinseveralexperiments.PreviousworkscarriedoutinourresearchlaboratorysuccessfullyshowedtheabilityofPAMtonon-invasivelymonitor 27

PAGE 28

tumorangiogenesisinvivo.Byusinganincidentlightwithanisosbesticwavelenght,thevascularestructurewasobservedintwomicebreastcancermodels[ 31 ].Wangetal.usedPAMtoimagethestructureofthevasculaturesurroundingamelanomainamouse.Theysuccessfullyimagedtheanatomyofthebloodvesselsandthemorphologicalrelationshipbetweenthemelanin-richtumorandthevasculature,byusinganisosbesticwavelenghtsatwhichmelaninandhaemoglobinhaveacomparablystrongabsorption[ 10 ].Thesamegroupperformedin-vivoexperimentsimagingthesubcutaneousmicrovasculatureofthepalmofahumanhandusingPAMaanisosbesticwavelenghtinthevisiblespectrum.Thetotalhaemoglobinconcentrationinthepalmwasimagedwithaqualitycomparabletothatintheaforementionedmousestudies[ 21 ].Inthisworkweusedphotoacousticmicroscopy(PAM)toimagethestructureofthevasculatureandalsotoevaluatehowthiskindofacquisitionismodiedbytheapplicationofanexternalpressureonthetargetregion.Apressuredevicewasdesignedandrealizedtoapplyanexternalpressuretotheinvestigatedregion.Firstly,weimagedthestructureofsubcutaneousvasculatureinthehumanwristwithPAMinabsenceofpressure.ThenweperformedPAMexperimentsonthesameareawhenanexternalpressurewasapplied.Inbothcases,wechose532nmlaserlight,whichisanisosbesticwavelengthforhaemoglobinanddeoxyhaemoglobin.Finally,theeffectsgeneratedbytheexternalpressurewereassessedintermsofchangesofowrateandbloodvolume,bymeasuringchangesintheamplitudeofthePAsignal.Infact,anincreaseinbloodowinacertainregionimpliesanincreaseinbloodvolumeinthesameregion.Ifbloodvolumeincreases,thenumberofchromophores,haemoglobinanddeoxyhaemoglobin,whichabsorblaserlightandgeneratePAsignal,grows.ThisimpliesanincreaseoftheamplitudeofthePAsignal.Ontheotherhand,adecreaseinthebloodowimpliesadecreaseinthevolumeofbloodinthesameregion.Ifthevolumedecreases,thenumberofchromophoresisreducedandtheamplitudeofthePAsignal 28

PAGE 29

decreases.Byconsequence,bymeasuringtheamplitudeofthegeneratedPAsignalitispossibletonoticeandassessthechangesoftheowrateandbloodvolume. 29

PAGE 30

CHAPTER2MATERIALSANDMETHODS 2.1IntroductionInthischapter,thematerialsandmethodsofourworkarepresented.Insection 2.2 hardwaresandsoftwaresusedinourexperimentsaredescribed.Theninsection 2.3 weshowandillustratethedesignofthephantomexperiments,carriedouttovalidatetheabsorption-basedpropertyofPAMimaging.SuchvalidationwouldenabletheuseofPAMforourin-vivostudytodetectvariationsinthenumberofchromophores(HbandHbO2),relatedtothevariationofbloodvolumecausedbytheapplicationofanexternalpressureonthetarget.Insection 2.4 thedesignofthepressureframe-baseddeviceisdescribed.Themostimportantissuesthatdroveusduringthedesignisillustrated.Thenwedescribethedesignofthedeviceandthewholesetupdesignedtoapplythepressure.Insection 2.5 weillustratethepreparationandexecutionofourin-vivoexperimentswithoutandwiththeapplicationoftheexternalpressure.Sinceduringin-vivoimagingofthesubcutaneousvasculaturenotonlythevesselsbutalsotheskingenerateaPAsignal,strategiestoremovetheskinsignalarenecessary.Twoapproachesaredescribedinsection 2.6 .TherstoneinvolvesthedevelopmentofaMatlabroutinetodetectandremovethesignalfromtheskin.Thesecondoneinvolvedtheuseofimagesegmentationtosemi-automaticallydetectandremovetheskinsignal. 2.2SetUp,DataAcquisitionandImageProcessingThehardwaresandsoftwarestoolsusedinourexperimentsareillustratedinthissection.Firstly,wedescribethesetupofthePAMsystem,usedtogeneratethelightanddeliverythistothesample.Thissystemalsocomprisesthetransducer,usedtodetectthePAwavesgeneratedbythesample,andadataacquisitionsystem.Secondly,thepostprocessingofthemeasuredsignalisdescribed,tounderstandhowthereconstructionoftheimagedvolumeisobtained. 30

PAGE 31

2.2.1PAMSetUpandPAMeasurementsThePAMequipment(seeFig. 2-1 )consistsofthreemajorsubsystems:atunablepulsedlasersystem,animagingheadmountedoncomputer-controlledmechanicalscannerandadataacquisitionsystem. Figure2-1. SchematicofthePAMsystem. Forthecoordinationofthesubsystems,asynchronizationsignalisrequired.Thedataacquisitionsystemreceivessynchronizationpulsesfromthelasersystemandrelaysthemtothemotorcontrollingsystem.Thelasersystemneedstoproducelaserpulsesshortenough(severalnanoseconds)inordertogenerateshortPApulsesthatprovidehighspatialresolution[ 23 ].Moreover,alasersystemwithhighpulserepetitionrateisdesiredforfastimageacquisition[ 26 ].ThelasersystemshouldalsobeabletosupplysufcientoutputenergytoensuregoodSNR[ 23 ],stillrespectingtheANSIsafetylimitforopticalenergy,bothinthevisibleandnearinfraredspectra[ 27 ].Inoursystem,ashort-pulsedlaserbeamof6nsdurationat20HzrepetitionrateisgeneratedbyaNd:YAGpulsedlasersystem.AsshowninFig. 2-1 ,thebeamisequallysplitbyanopticalsplitterintotwobeams.Alensfocuseseachbeamintoaberbundle,inwhich 31

PAGE 32

thelightcanpropagateefcientlyuntiltheultrasoundproberightoverthesample.Theimagingprobeconsistsoftheacoustictransducerandthetwoberbundlesandcanbemovedalongthex)]TJ /F2 11.955 Tf 12.64 0 Td[(y)]TJ /F2 11.955 Tf 12.64 0 Td[(zaxisbyamotorizedstagecontrolledwithaLabviewinterface.Theheightandthehorizontalpositionoftheimagingheadhavetobeadjustedtomatchtheregionofinterest(ROI).Theimagingheadisimmersedinawatertankwithawindowsealedwithanopticallyandultrasonicallytransparentmembrane,toreducetheacousticalimpedanceandoptimizethetransmissionoftheultrasoundswavestothetransducer.ThePAMset-upcomprisesalsoawidebandpreampliertoamplifythedetectedPAsignals.ThesignalisthendigitizedandstoredinaPC.TheholdingtimeinthedataacquisitionprogramhastobesetsothatenoughtimeisallocatedbetweenBscansforthetranslationstagetoreturntoitsinitialyposition.Duringthescanning,thedataacquisitionPCstoresthedigitizedPAsignals.WhenoneBscaniscompleted,thedataacquisitionPCblocksthesynchronizationpulsesfromthemotor-controllingsystemduringtheholdingtime.AsshowninFig. 2-2 ,theA-lineisaone-dimensionaldepth-resolvedimagealongthezaxisconvertedfromthetime-resolvedPAsignalsrecordedateachhorizontalposition(x)]TJ /F2 11.955 Tf 11.96 0 Td[(yplane). Figure2-2. ImageformationinPAM.Figureadaptedfrom[ 29 ]. TherecordingdurationofthePAsignalisdeterminedbythedesireddepthrangealongthezaxisandthesoundvelocityvinsofttissue(v=1.54mm/s).Thenthe 32

PAGE 33

motorizedstagetranslatestheimagingheadbyonescanningstepalongthexaxisandawaitsthenextlaserpulse.Oncethelateralscanalongthexaxisiscompleted,atwo-dimensionalcrossectionalimage,i.e.aBline,isobtainedandtheimagingheadisreturnedtoitsinitialxcoordinate.Then,itisfurthertranslatedbyonescanningstepalongtheyaxistostartanewBscan.Acompleteraster-scanningalongthex)]TJ /F2 11.955 Tf 11.72 0 Td[(yplaneproducesthenalvolumetricimage. 2.2.2PostProcessingoftheMeasuredPASignalThestepsofthesignalprocessingisschematizedinFig. 2-3 andexplainedinthisparagraph.AfterthePAmeasurementiscarriedout,thedataobtainedareprocessedwithMatlab.Matlabreceivesavectorofnumericaldataasinput.ThevectorisasequenceofseveralB-scan,acquiredasdescribedabove.Foreachelementofthesequence,themeanvalueissubtractedandtheHilberttransformiscomputed.Then,thenewsequenceisstoredina3-dimensionalarray,insuchawaythateachdimensionofthetensorcorrespondstoadimensioninthetomographicspace.Foreachtomographicplane,theabsolutevalueiscomputed.Afteraminimumandamaximumvalueshadbeenfoundamongtensor'svalues,thedataarenormalizedinto[0-255]grayscalerange.Finally,Matlabgeneratesasetofimagesasoutput:eachimagerepresentsatomographicplaningrayscale.TheseimagesareloadedinAmiraafterthevoxelsizeisset,theimagedvolumeisreconstructed.Adigitalmedianlterisappliedtothedata,toreducethesaltandpeppernoise.FinallytheMaximumIntensityProjection(MIP)ontheorthogonalplaneisdisplayed. 33

PAGE 34

Figure2-3. Flowchartoftheexperimentalprocedure,fromthemeasurementofthePAsignaltothereconstructionoftheimagedvolume. 2.3PATheoryValidationwithPhantomExperimentsThepurposeoftheseexperimentsistovalidatethePAMtheory,accordingtowhichthecontrastisproportionaltotheopticalabsorptionofthetarget(see 1.1.1 ).Particularly,weaimtoprovethatanincreaseofthenumberofchromophoreswhichabsorbthelightisrelatedtoanincreaseoftheamplitudeofthemeasuredPAsignal,andviceversa.Theincrease,ordecrease,ofsuchamplitudecanbequantiedbycalculatingthecontrast. 34

PAGE 35

2.3.1PhantomPreparationAsetofpreliminaryexperimentswascarriedoutbyusingagarosephantoms.Agaroseisaplant-basedjellifyingagent.Itissoldintheformofawhitepowderthat,oncemixedwithwater,heatedabove85Candthencooledtobelowaround35C,becomesalightlyopaquegelatinoussubstance.Agaroseisusedforitstissue-equivalentspeedofsoundandbecauseithasmechanicalandopticalfeaturessimilartothoseofsofttissues.Itsmainlimitationisthelowtoughness,whichmakesitfragileduringhandling.Theroleofagaroseistoholdthetargetobjectinplacewithinthephantom.Inourphantoms,thetargetismadeofaspecicamountofinkaddedtothesameagarsolutionthanthebackground.Byvaryingtheamountofinkaddedtothesolution,differenttargetsarerealized,eachwithaspecicabsorptioncoefcienta.Inparticular,themoreinkisadded,thehighertheais.OneofourphantomsisshowninFig. 2-4 Figure2-4. Tumormimicphantom.Thetargetisthedarkspotlocatedinthecenterofthephantom.Thetargetissurroundedbytheagarosebackground. Weembeddedthreedifferent2mm-diameterroundedtargets,5mmunderthephantomsurface.Theabsorptioncoefcientofthebackgroungis0.007mm)]TJ /F6 7.97 Tf 6.58 0 Td[(1,whiletheabsorptioncoefcientsofthethreetargetsis0.021mm)]TJ /F6 7.97 Tf 6.59 0 Td[(1,0.049mm)]TJ /F6 7.97 Tf 6.59 0 Td[(1and0.07 35

PAGE 36

mm)]TJ /F6 7.97 Tf 6.58 0 Td[(1,whichare3times,7timesand10timeshigherthanthebackground,respectively.However,thereducedscatteringcoefcientofthebackgroundandthetargetsis1.0mm)]TJ /F6 7.97 Tf 6.58 0 Td[(1. 2.3.2PhantomExperimentsAlow-frequency(3.5MHz)transducer,with3.5mmfocallength,wasusedtodetectthephotoacousticwaves.Ascanningstepsizeof75mandasquare-shapedimagingareawitha75mmsidewereset.Thelaserwavelenghtwastunedto710nm.ThiswavelenghtintheNIRspectrumwaschoseninsteadofawavalenghtinthevisiblespectrumtoreducetheeffectofthescatteringphenomenon,sincethescatteredintensityisproportionalto1=4[ 35 ].Theaveragevaluesofcontrastwerecomparedamongthethreesetsofdata.Thecontrastisadimensionlessnumber,calculatedasexpressedbytheEqn. 2 :contrast=AROI)]TJ /F2 11.955 Tf 11.96 0 Td[(Abackground Abackground. (2)whereAROIisthearithmeticmeanofthenormalizedamplitudesofthesignalincorrespondencetotheregionofinterest(ROI),whichistheregionwherethetargetislocated;Abackgroundisthearithmeticmeanofthenormalizedamplitudesofthesignalincorrespondencetothebackground. 2.4DesignofthePressureDeviceAspreviouslyexplained,theaimofthisstudyistoin-vivoimagethestructureofthesubcutaneousvasculatureandtoevaluatehowsuchkindofimagingismodiedbytheapplicationofanexternalpressureonthetarget.Weneededatailoreddevicethatwasincontacttothewristsurfaceandexertedamechanicalpressureonthetissue.Therefore,apressureframe-baseddevicewasdesignedforthispurpose. 2.4.1MotivationsoftheDesignTopologyTofullyunderstandthecriteriathatwereconsideredwhiledesigningthedevice,itisimportanttoemphasizesomefeaturesofthesetupwherethedeviceisplaced. 36

PAGE 37

Thepressuredevicehastocloselysurroundthetargetwithouttouchingtheprobewheretransducerandopticalbersaremounted.Thisisthemostimportantissuewhendesigningthedevice,becausetheprobeisveryclosetothetargetsurface.Infact,asalreadyexplainedinsection 2.2 ,thetransducerhasafocaldistanceof6mm,whichmeansthatprobehastobexedabout3mmovertheskinsurfacetofocusthebloodvessels.Thus,theavailablespacetopositionthepressuredeviceislimited.AnotherissueinvolvesthetransmissionofthePAwavesfromthetargettothetransducer.Infact,ifsomethingisinbetween,thePAwaveswouldbeattenuated.Therefore,thepartofthedevicewhichisincontactwiththetargetdoesnothavetocovertheimagingarea,nottoimpedethePAwavestransmission.Moreover,asdescribedinsection 2.2 ,theprobeismovedalongthex)]TJ /F2 11.955 Tf 12.19 0 Td[(y)]TJ /F2 11.955 Tf 12.2 0 Td[(zaxisbyamotorizedstagetorasterscantheimagingarea.Thus,thepressuredevicedoesnothavetoimpedethemovementsoftheprobewhileexertingthepressure.Fromtheseconsiderations,itemergesthatthemostcriticalissuesinvolvethedesignofthepartofthepressuredevicewhichdirectlyexertsthepressureonthetarget.Forthisreason,aring-shapedpartwasdesigned.Withthisdesign,theringcanbepositionedoverthetargetandtheprobecanbeplacedinsidethering.Someconsiderationshavetobedoneaboutthechoiceoftheringsize.Thesmallertheringis,thegreaterthechangesofvesselvolumeareintheregioninsidetheringduetothepressure.Ifthepressurewasexertedthroughapartwhosedimensionsarecomparablewiththeimagedarea,theeffectofthepressureonthebloodowwouldbemuchmoreappreciable.Ontheotherhand,theringhastobelargeenough,becausetheprobehastobeplacedinsideitandithastomovetoscantheimagingareaalongthex)]TJ /F2 11.955 Tf 11.96 0 Td[(yplan.Someimportantconsiderationsaboutthepressureapplicationhavetoberaisedheretounderstandhowthepressureapplicationconditionedthedesignofthedevice.Wechosetoapplythepressurebyimposingaloadonthedevice,sothattheforceisappliedbytheringonthetarget.Wedecidedtousesomeweights,astheoneshown 37

PAGE 38

Figure2-5. Oneofthecalibrationweightsusedtoapplyaloadonthedevice,exertingapressureonthetarget.Photocourtesyofauthor. inFig. 2-5 .Particularly,severalweightsof100gand500ghavebeenchosenforourpurpose.Thesizeoftheformeris2cmby2cmby4cm;thesizeofthelatteris4cmby4cmby6cm. 2.4.2TechnicalDrawingAsquaredring-shapedpartwithatcontactsurface(i.e.,witharectangularsection)wasdesignedtodirectlyexertthepressure:theringispositionedonthetargetandtheprobeisplacedinsidethering.Accordingtothisarrangement,thePAwavesarereceivedbythetransducerwithoutbeingattenuatedandtheproblemofthelimiteddistancebetweenthetransducerandthetargetisovercome.Theringisplacedinthewatertank,overthewindowwhichissealedwiththemembrane,incontactwiththetarget.Theprobehasadiameterof2.5cmandithastomovetoscananimagingareaof9mmby9mm.Hence,theringwedesigned,showninFig. 2-6 ,hastheinternalsideof3.7cm,theexternalsideof4.7cmtoallowthemovementoftheprobeinthex)]TJ /F2 11.955 Tf 12.45 0 Td[(yplane.Thethicknessis0.5cm.Sincethereisnotspaceavailabletoplacetheweightsrightoverthering,aretainingstructurewasdesignedtosustainthem.Thisstructureisconnectedwiththeringandisverticallydeveloped.Itismadeupofthreemajorparts:abottompartwhichcontainsthering,atoppartwhichholdstheweightsandfourverticalrodstoconnectthetopandthebottomparts.Fourbranchesdepartfromthering.Thesebranchesand 38

PAGE 39

Figure2-6. Squaredring-shapedpartofthepressuredevicewhichexertsthepressuredirectyonthetarget.Theinternalsideis3.7cm,theexternalsideis4.7cm. Figure2-7. Frontviewofthemodelofthebottompart. theringformthebottompart,whichisanuniquepartshowninFig. 2-7 andinFig. 2-8 .Alengthof4.5cmwaschosenforeachbranch,sothattheentirepartwassmallenoughtobeplacedinsidethewatertank.Eachbranchliesonaplanewhichis30angledwithrespecttotheplanewheretheringlieson.Thankstothisconguration,onlythering 39

PAGE 40

Figure2-8. Sideviewofthemodelofthebottompart. Figure2-9. Frontviewofthemodelofthetoppart. isincontactwiththetargetandtheotherpiecesofthestructuredonottouchanyotherpartofthesetup.Therefore,allthepressureisexertedtothetarget.Thewholeparthasalengthof12cmandawidthof7cm.Theextremityofeachbranchhasathreadedhole,designedtoxtherestoftheretainingstructure.Particularly,averticalrodisscrewedineachofthem.Eachrodhasalengthof12cmandaroundsection,withadiameterof0.5cm.Theotherextremitiesoftherodsaretightenedthroughsomescrewstothetoppart,whichisshowning. 2-9 .Thetopparthasanx-shapewithtwoholesinside,toreducethebulkofthestructure.Thelengthandwidthofthispartperfectlytwiththelengthandwidthofthebottompart.ThemodelofthewholestructureisshowninFig. 2-10 .Thebottompartandthe 40

PAGE 41

Figure2-10. Modelofthewholestructure:thetoppartandthebottompartarexedtogetherandconnectedbythefourverticalcylinders. toppartwasprintedbyusinga3Dprinter.Thematerialisahard,rigid,semi-transparentresin.Theuseofametallicmaterialwasdiscardedbecausemetallicsurfacesmyabsorbscatteredlaserlightandgeneratespuriousphotoacousticsignal[ 54 ],whichdoesnothappenwithplasticmaterials.Thefourcylindricalrodshadalreadybeenpurchasedbyourlab,thustheywerenotprintedbythe3Dprinter.Theirmaterialisstainlesssteel.ApictureofthewholestructureisshowedinFig. 2-11 2.4.3SetUpDevelopedforthePressureApplicationThewaythepressuredeviceisplacedintheexperimentalset-uptheandthenusedisshowninFig. 2-12 .Thebottompartofthepressuredeviceisimmersedinthewatertank(indicatedbythelettera),incontactwiththetransparentmembrane.The 41

PAGE 42

Figure2-11. Pictureofthewholestructure:thetoppartandthebottompartaremadeofasemi-transparentresinwithayellowishtintandarexedtogetherbythefourverticalcylindersmadeofsteel.Photocourtesyofauthor. probecontainingtransducerandopticbers(b)ispositionedbetweenthebottompartandthetoppartofthepressuredevice.Aplasticcontainer,indicatedbytheletterc,isxedonthetopofthestructure,tosafelyholdtheweightsonthepressuredevice.Byknowingtheforce,itispossibletocalculatethepressuredividingbytheareaoftheringcontactsurface.Thepartsd,eandftogethersustainthepressuredevice.Particularly,dande2aretwometallicrodsrmlyxedtogetheratright-angle.Thetworodstogetherarelockedwithalinearstage(f),whichisrmlyxedtothetable.Byrotatingthethreehandlesofthelinearstage,thepositionofthetworodsandsothepositionofthewholepressuredevicecanbepreciselymanuallyadjustedbeforestartingeachexperiment.Aspring-dynamometer,xedtothehorizontalrod(d)and 42

PAGE 43

indicatedbytheletterg,isusedtoestimatethereactionforceexertedbythetissuewhenthepressureisapplied.Itcontainsaspringthatmaybecompressedorpulledapartinthedirectionofitslongeraxisanditcalculatestheforcebyknowingthespringconstantandmeasuringtheextensionorcompressionofthespring.Thespringinsidethedynamometerisconnectedwithanexternalhook,fromwhichthepressuredeviceishungup.Thedynamometerisalsousedinthecalibrationstage.Inthisstage,noweightisplacedintheplasticcontainer,sotheonlyactingforceistheweightofthepressuredevicehungbelowthedynamometer.Theverticalpositionofthepressuredeviceissetdownstepbystepuntiltheringtouchesthetargetwithoutexertinganyactualpressureonit.Inthiscondition,thedynamometeroffsetisadjustedtozeroandthisisconsideredthereference-condition.Afterthecalibrationstage,thedeviceisreadytouse.Whentheweightsareplacedonthepressuredevice,theactualforceexertedisrelatedtothereference-conditionbysubtractingtheincreaseoftheforcemeasuredbythedynamometerduetodownwarddeformationtotheaddedweight. 2.5InVivoPAMExperiments:ImagingoftheSubcutaneousVasculatureSeveralin-vivoexperimentswereperformedtoimagethesubcutaneousvasculatureofthehumanwristunderdifferentexternalpressures.First,thesubcutaneousvasculaturewasimagedwithouttheapplicationofanyexternalpressure.Then,otherPAacquisitionofthesameareawerecarriedoutwhenthepressuredevicedescribedinsection 2.4 wasexertingthepressure.Thesetupwasuseddescribedinsection 2.2 .Fortheseexperiments,thelaserwastunedtoafrequencyof532nm,anisosbesticpointforhaemoglobinanddeoxyhaemoglobininthevisiblespectrum.Thehighcontrastofhemoglobinat532nmenablestheuseoflow-levellaserexposure.Infact,beamsplitter,mirror,lensesandberbundleswereopportunelycoupledsothatthepowerofthelightcomingfrombothberbundlestothesamplewas40mW.Astherepetitionrateofthelaseris20Hz,theenergyofeachpulsewas2mJ.Thesizeofthelightspotattheopticalfocalpointwas 43

PAGE 44

Figure2-12. Set-uparrangedtoexertthepressureonthetarget.(a)Watertankwherethepressuredeviceisplaced.(b)Photoacousticprobe,wherethetransducerandopticalbersareinserted.(c)Plasticcontainertosafelyholdtheweightsusedtoloadthestructure.(d,e)Metalliccylindricalrodswhichsustainthepressuredevice.(f)Linearstageusedtoadjustthepositionofthewholestructurexedtoit.(g)Dynamometerusedtomeasuretheappliedforce.Photocourtesyofauthor. 0.2cm2,thustheexposurebyanysinglelaserpulseattheopticalfocalpointwas10mJ=cm2,whichisundertheANSIlasersafetylimit(20mJ=cm2).Atransducerwith50MHzascentralfrequency,with3mmapertureand6mmfocallengthwasused.Ityields15mresolutioninaxialand60mresolutioninlateralatthefocalpoint.Thescanningstepsizewassetto60mandanimagingareaof6mmby9mmwaschosen.The 44

PAGE 45

imagingdepthis3mm.Withtheseparameters,oneB-scantakes8sandtheimagingofthewholeareatakesabout15min.Foreachimagingarea,theacquisitionwasperformedvetimesinarow,eachtimeunderadifferentpressurecondition.Thusvedatasetswerecollectedforeachsetofexperiments.Werstimagedthehumanwristvasculaturewithoutapplyinganyexternalpressure.Thepressuredevicewasplacedintheset-upstartingfromtherstexperiment,asshowninFig. 2-13 .Ifthisexpedienthadnotbeenadopted,theprobewouldhavebeenmovedaftertherstexperimenttoinsertthepressuredeviceinthewatertankforthefollowingexperiments.Thishadtobeavoided,becauseachangeintheprobeposition,meansachangeoftheimagingareabetweenoneacquisitionandthefollowingone.Fig. 2-13 showsthepositionofthepatientandofthepressuredeviceduringtherstexperiment.Thewristandtheelbowofthepatientwerepositionedontwoadjustableholders,tomakesurethatthepatientwasinacomfortableposition.Theimagingareaofthewristwascoveredwiththeultrasoundgelandplacedunderthewatertank,incontactwiththeplasticlm.Theringofthepressuredevicewaspositionedinthex)]TJ /F2 11.955 Tf 12.99 0 Td[(yplanesothatthecenteroftheringwascoincidentwiththeimagingarea.Duringthisstage,thepressuredevicewasnotloadedbytheweightsandtheringwasnotincontactwiththetarget,thusnopressurewasexertedonthetarget.Aftertheacquisitionoftherstdatasetwascompleted,thepressuredevicewasmanuallysetdownstepbystepusingthelinearstage,untiltheringtouchedthewrist.Atthispointthedynamometermeasured0N.Afterwards,fouracquisitionswerecarriedouttoimagethesameareaunderfourdifferentpressurelevels:7000N,15000N,20000Nand25000Nrespectively.Eachacquisitionwasstartedrightaftertheendofthepreviousone.25000Nwaspreviouslyidentiedasthemaximumloadthatthepatientcantoleratewithoutreportinganyunpleasantorpainfulsensationcausedbythepressingring.Theothervalueswere 45

PAGE 46

Figure2-13. Positionofthepatientandofthepressuredeviceduringtherstexperiment.Anypressurewasexertedonthetarget.Photocourtesyofauthor. chosenbypickingfourvaluesasmuchaspossibleequallydistributedinthespan0-25000N.TheloadwaspositionedintheplasticcontaineronthepressuredeviceasshowninFig. 2-14 .Oncethepressureisappliedonthetarget,thewristsurfaceundergoesasmallverticaldisplacement.Forthisreason,ineachexperimentafterthepressureapplication,theverticalpositionoftheprobewasadjustedbeforestartingtheacquisition,tomakesurethatthebloodvesselswereproperlylocatedonthefocalplan.Thisexperiment,consistingofveacquisitionsforthesameimagingareaunderdifferentpressurelevels,wasrepeatedfourtimes.Allthedatasetswereprocessedasdescribedinsection 2.2 .Asetofimagescorrespondingtothetomographicplanes,was 46

PAGE 47

Figure2-14. Positionofthepatientandofthepressuredevicewhenthepressureisexertedonthetargetregion.Theweightsareplacedintheplasticcontainerxedonthetopofthepressuredevice.Photocourtesyofauthor. generatedwithMatlab.ThenAmirawasusedtoreconstructthevolumeandremovethenoise.Unliketheotherexperiments,inthiscaseAmirawasnotusedtocomputeanddisplaytheMIP.Rather,thereconstructedvolumewasexportedontoMatlab,becausetheanalysisoftheamplitudeofthesignalbyusingMatlabispreferred.Finally,MatlabwasalsousedtocomputeanddisplaytheMIP. 2.6RemovaloftheSignalFromtheSkinTheskinisamelanin-richstructure.BothinthevisibleandintheNIRrange,theabsorptioncoefcientofmelaninishigherthanthoseofhaemoglobinanddeoxyhaemoglobin.Inparticular,ascanbeseeninFig. 1-2 ,attheNIRwavelengths,theabsorptioncoefcientofmelanincanbeevenoneorderofmagnitudehigherthantheabsorption 47

PAGE 48

coefcientofthetwospeciesofhaemoglobin.Thus,melaninrepresentsanimportantsourceofcontrastinPAimaginganditsPAsignalmightovershadowtheweakersignalfromthesubcutaneousvesselsinsomeA-lines[ 22 29 ].Moreover,becausetheunevendistributionofmelaninpigmentsintheskin,thePAamplitudesgeneratedfromtheskinsurfacevarywithhorizontalpositions.Fig. 2-15 showstheMIPofalltheB-scansalongthex)]TJ /F2 11.955 Tf 12.38 0 Td[(zplan.Twolayersareclearlyvisible.Thelayeronthetoprepresentstheskin,whilethebottomlayerrepresentsthesubcutaneousbloodvessels.Ascanbeseenfromthecolorsing. 2-15 ,theamplitudeofthesignalfromtheskinishigherthanthatfromthebloodvessels.Moreover,thepicturealsoshowsthattheskinlayermaynotbeatandthatthemelanindistributionisunevenanddiscontinuousinsomeparts. Figure2-15. MIPalongthex)]TJ /F2 11.955 Tf 11.95 0 Td[(zplan.Thetoplayerrepresentstheskin,whilethebottomlayerrepresentsthesubcutaneousbloodvessels. IftheMIPalongthex)]TJ /F2 11.955 Tf 12.5 0 Td[(yplaniscalculateddirectly,PAamplitudesfromtheskinsurfaceratherthanthosefromthevesselsareprojectedontothenal2Dimage.Thisresultsinerrorsinthevisualizationofthetruevasculature,asshowninFig. 2-16 48

PAGE 49

Figure2-16. MIPalongthex)]TJ /F2 11.955 Tf 11.95 0 Td[(yplan.Theskinsignalisprojectedontothenalimage,makingthevasculaturebarelyvisible. Therefore,detectingtheskinproleandsubsequentlyremovingtheskinsignalsaretwoimportantsteps,beforetheMIPimagesarecalculated. 2.6.1AutomaticAlgorithmforSkinProleDetectionandRemovalAMatlabprogramwasdevelopedtodetecttheskinprole,removeitfromthevolumetricdataandcalculatetheMIPimages.TherststepfortheskinproledetectionwascarriedoutanalyzingthefeaturesofeachA-line.Basingontheanalysisoftherelationshipbetweentheamplitudeofthesignalgeneratedfromtheskin(As)andthesubcutaneousbloodvessels(Av)ineachA-line,aroughestimationoftheskinlocationwasperformed.Withinthedepthrange,mostA-lineshaveonlytwodominatingPApeaks,whicharegeneratedfromtheskinsurfaceandasubcutaneousvessel,respectively.TherelationshipbetweentheamplitudeofPAsignalgeneratedfromtheskinsurfaceandtheamplitudeofPAsignalgeneratedfromasubcutaneousbloodvesselcanbeclassiedintothreemajorcategorizes,asFig. 2-17 shows:InFig. 2-17 A,AsandAv 49

PAGE 50

Figure2-17. Relationshipbetweenamplitudesofthesignalfromskinandbloodvessels,AsandAvrespectively.A)AsandAvarecomparablystrong.B)Asdominates.C)Avdominates.Figureadaptedfrom[ 29 ]. arecomparablystrong,inFig. 2-17 BAsdominateswhileinFig. 2-17 CAvdominates.However,othersituationsalsoexist.Insomecases,morethantwodominatingPApeaksexistwithinanA-linewhen,forexample,multiplePApeakscanbegeneratedfrommultiplebloodvesselsatdifferentdepthsatthesamehorizontalposition.Inothercases,anA-linecouldshowonlyoneornodominatingPApeak,i.e.,itonlyhasaPApeakgeneratedfromtheskinsurfaceoritconsistsonlyofrandomnoiseandbackgroundsignalswithoutdistinctivePApeaksfromeithertheskinsurfaceorbloodvessels[ 29 ].AllthesesituationsaregeneralizedtohaveNpeaksintherst-steproughestimationforthe1-DskinproleineachB-scanimage.Nisavaryingnumberthatcanbeadjustedaccordingtoeachspecialsituation.InB-scanimages,A-linesareinanorderthatfollowsthedataacquisitionsequence.ForeachA-line,thedepthlocationsoftherstNstrongestPApeaks(Pi,i2[1,N])areidentiedandthedistancebetweeneachpairofconsecutivepeaks(Dj=Pj+1)]TJ /F2 11.955 Tf 12.59 0 Td[(Pj,j2[1,N)]TJ /F8 11.955 Tf 12.56 0 Td[(1])arefurthercalculated.Oncethelargestpairdistance(Djmax=maxDj)isfound,PjmaxisconsideredthelocationoftheskinsurfaceinthatA-line.SuchanoperationisbasedontheobservationthatthedistancebetweenthePApeakgeneratedfromtheskinsurfaceandthePApeakgeneratedfromashallowerbloodvesselisgenerallylargerthanthedistancesbetweenanyotherconsecutivepairswithinthesameA-line.AfterallA-lineswithinaB-scanimageareprocessedasdescribedabove,a1-Droughextimationoftheskinproleisacquired. 50

PAGE 51

Thesecondstepfortheskinproledetectionwasachievedbyremovingtheerrorpoints,oroutliers,throughsmoothing.Theassumptionbehindthesmoothingoperationisthatthemajorityoftheestimatedskinlocationsintherst-roundroughestimationwereeithertrueorwereclosetotheirtrueskinlocations,andtheoutlierswererandomlydistributedalongthexaxis[ 29 ].Hence,amethodisneededthatcanextractthebackbonebasedontheestimatedmajorskinlocationsandthatwillnotbeaffectedbytheoutliers.Moreover,suchamethodshouldnotrequirehumanintervention.Thenonparametricrobustlocallyweighted-regressionsmoothingwasadopted.Comparedwithparametricsmoothing,nonparametricsmoothingoffersbetterapproximationsthanparametricsmoothingwhenappliedtotheirregularlyshapedskinproleAfterthe1-DsmoothedskindetectionwascompletedforeachB-scanimage,thesamesmoothingprocedurewasperformedagainalongtheyaxisandthenalskinprolewasobtained.Finally,thecomplete2-Dskinprolewasacquiredbyapplyinga2-Dmedianspatiallteringtosmooththeimage,reducingthenoisebutstillpreservingdiscontinuities.Thelaststepconsistsoftheremovaloftheskinsignalfromthevolumetricdata.Dependingonthedistributionoftheskinsignalanditsamplitude,post-processingthedatawiththeautomaticalgorithmmightbeenoughtoensureagoodimagequality.ForexampleinFig. 2-17 a,whenAsandAdarecomparable,theroughextimationgivesthecorrectskinlocation.SimilarlyinFig. 2-17 b,whenAsdominates,thesecondstrongestpeakthatthealgorithmselectsisneartherealskinsurface.However,inFig. 2-17 c,whereAvdominates,thesecondstrongestpeakisthepeakfromthebloodvessels,sothealgorithmgaveanerroneousestimationoftheskinlocationinthiscase.Hence,partoftheskinsignalmightbestillpresentintheMIP,affectingthequalityofthenalresult.Fig. 2-18 showstheMIPimageofthesamedatasetofFig. 2-16 aftertheskinremovalbytheautomaticalgorithmimplemented.Thegreatpartoftheskinsignalwasremovedbythealgorithmandmostofthesubcutaneousbloodvesselsarevisible.Howevertheskinsignalisstillpresentinthebottom-leftcorner,withacloudofpointspattern. 51

PAGE 52

Figure2-18. MIPalongthex)]TJ /F2 11.955 Tf 11.95 0 Td[(yplanaftertheskinremovalbyusingtheautomaticalgorithmimplemented. Inthesepoints,theroughestimationgaveanerrorandtheskinsignalhasbeenonlypartiallyremoved. 2.6.2ImageSegmentationandRemovaloftheSkinSignalAnotherapproachwastakenintoaccounttoremovetheskinsignalwhentheautomaticalgorithmdidnotgivethecorrectskinlocation.Thestrategyweadoptedinvolvesimagesegmentationbymeansofthesnakeregularizedborderdetectionalgorithm.ThesoftwareITK-SNAPwasusedforthispurpose.Asrst,thevolumecontainingthedatainaNIfTI-1DataFormatisloadedontoITK-SNAP.Bydefault,ITK-SNAPassignsthespeciallabel0toeachvoxelinthevolume:thislabelmeansthatavoxelhasnotbeensegmentedyet.ThemethodologybehindITK-SNAPiscalledsnakeevolution.Thetermsnakeisusedtorefertoaclosedcurve(orsurfacein3D)thatrepresentsasegmentation.Insnakeevolutionmethods,thesnakeevolvesfromaveryroughestimateoftheanatomicalstructureofinteresttoaverycloseapproximationofthestructure.Therststepofthesegmentationprocedureistheinitialization.Itconsistsofmanuallyplace 52

PAGE 53

Figure2-19. Sectionoftheimagedvolume.Theskinproleisshowninthewhitebottomlayer.Theotherwhitespotsovertheskinrepresentpartsofbloodvessels.Someseedpoints,red-colored,havebeenplacedalongtheskinprole. seedpointsalongthecontouroftheskin,slicebyslice.Theseseedpointsgrowovertimetotakeshapeoftheskin.Thesnakeevolutionisgovernedbyamathematicalequationthatdescribesthevelocityofeverypointonthesnakeatanyparticulartime.Thevelocityofeachpointdependsonthedistributionoftheseedpointsandontheintensitiesoftheimageintheneighborhoodofthepoints.Thevelocitiesaregreaterintheregionsoftheimagewheretheintensityishomogeneousandweakerwheretherearediscontinuitiesinimageintensity.Fig. 2-19 showsasectionofthevolume,wherebloodvesselsandtheskinaredistinguishable.Theskinisthecontinuouswhitebottomlayer,whiletheotherwhitespotsrepresentpartsofbloodvesselsinthesamesection.Someseedpointwerenplacedalongtheskinprole.Fig. 2-20 showsthesamesectionofFig. 2-19 aftertheseedpointsgrewandtooktheshapeoftheskin.Thered-coloredvoxelscorrespondingtotheskinsurfaceareassociatedwiththelabel1.However,thesnakeevolutionmightnotlocatesomevoxelsthatactuallyrepresenttheskin,becausetheskinprolemaybediscontinuous.ThiscanbeseeninFig. 2-20 ,wheretheskincontourwasnotfullylocated,sincetheleftpartoftheproleisstillwhiteandnotassociatedtotheskinlabel.Toxthisproblem,thosevoxelsweremanuallypaintedandmarkedwiththeskinlabel,slicebyslice. 53

PAGE 54

Figure2-20. Sectionoftheimagedvolume.Theseedpointsgrewandtooktheshapeoftheskin.Theleftportionoftheskinproleisnotcontinuous,thusithasnotbeenlocatedbythesnakeevoutionalgorithm.Thisproblemhasbeensolvedbymanuallypaintingthosevoxels. Theresultofthesegmentationisavolumewiththesamedimensionsoftheoriginalvolume,inwhicheachvoxelwasassociatedto1or0,accordingasthatvoxelispartoftheskinornotrespectively.Byusingthisvolumeasabinarymask,withaMatlabroutinethatweimplemented,theskinsignalcanberemovedfromtheoriginalvolume.Fig. 2-21 showstheMIPimageofthesamedatasetofFig. 2-16 aftertheskinwassegmentedandremoved.Thesubcutaneousbloodvesselsarecompletelyvisible.Thisshowsthatthesegmentationisausefultooltoenhancetheimagequalitybyremovingtheskinsignal;howeverthisprocedureistimeconsuming,sincetheaccuratesegmentationofavolumetakesabout1hourandhalf. 54

PAGE 55

Figure2-21. MIPalongthex)]TJ /F2 11.955 Tf 11.95 0 Td[(yplanafterthesegmentationandtheskinremoval. 55

PAGE 56

CHAPTER3RESULTS 3.1IntroductionInthischaptertheresultsofourexperimentsareillustrated.First,theresultsofthephantomexperimentstovalidatetheabsorption-basedpropertiesoftheimagingareshown.Theseresultscanbereadandevaluatedbothqualitatively,bycomparingtheMIPs,andquantitatively,bycomparingthecontrast.Secondly,theresultsofthein-vivoexperimentsareillustrated.TheMIPsprovideacomparisonbetweenthecasesunderdifferentconditionsofexternalpressure.However,amoreaccuratecomparisonisperformedbyevaluatingchangesinamplitudeofthesignalduetothepressureapplication. 3.2PATheoryValidationwithPhantomExperimentsTheresultsofourtumor-mimicphantomexperimentsareshowninthissection.Fig. 3-1 ,Fig. 3-2 andFig. 3-3 representthemaximumintensityprojection(MIP)imagesofthephotoacousticsignalsprojectedontheorthogonal(x-y)plane.Theamplitudeisrepresentedinacolormaprangingfromred(lowamplitude)towhite(highamplitude).Fig. 3-1 istheMIPofthephotoacousticsignalgeneratedwhentheabsorptioncoefcientofthetargetisequalto0.07mm)]TJ /F6 7.97 Tf 6.59 0 Td[(1,whichis10timeshigherthantheabsorptioncoefcientofthebackground.Fig. 3-2 istheMIPofthephotoacousticsignalgeneratedwhentheabsorptioncoefcientofthetargetisequalto0.049mm)]TJ /F6 7.97 Tf 6.59 0 Td[(1,7timeshigherthantheabsorptioncoefcientofthebackground.Fig. 3-3 istheMIPofthephotoacousticsignalgeneratedwhentheabsorptioncoefcientofthetargetisequalto0.021mm)]TJ /F6 7.97 Tf 6.59 0 Td[(1,whichisonly3timeshigherthantheabsorptioncoefcientofthebackground.ThetargetisshowninthecenterofFig. 3-1 ,Fig. 3-2 andFig. 3-3 .TheotherbrightspotinthecornerofeachframerepresentsthePAsignalgeneratedbyareferencespot,takenasstartingpointtolaunchtheacquisition.Itisnotpartofthe 56

PAGE 57

phantomanddonothavetobeconsideredtointerprettheresults.Asstatedinsection 2.3.1 ,asquare-shapedimagingareawitha75mmsidehasbeenchosen.ItisinterestingtocomparethethreeresultsintermsofamplitudesofthePAsignal,bycalculatingthecontrast.ThevalueofcontrastwascomputedforeachresultasexpressedbyEqn. 2 ,toquantitativelyassessthecontrastintheseresults.ThetargetregioninFig. 3-1 presentsthehighestvalueofabsorptioncoefcienta,duetothehighestquantityofinkaddedtothesolution.Becauseofthehigha,theamplitudeofthephotoacousticsignalishigh,asthecenterofFig. 3-1 shows.Theseobservationsarequantitativelyconrmedbythehighaveragevalueofthecontrast:12.8159.Onthecontrary,thetargetregioninFig. 3-3 isbarelyvisible,becausetheabsorptioncoefcientofthetargetisonlythreetimeshigherthantheabsorptioncoefcientofthebackground.Thisqualitativeobservationisconrmedbytheaveragevalueofthecontrastcomputedforthissetofdata:1.1724.Intheend,theresultsinFig. 3-2 showthattheamplitudeofthesignalfromthetargetishigherthanthecaseofabsorptioncoefcientequalto0.021mm)]TJ /F6 7.97 Tf 6.59 0 Td[(1,butlowerthantheresultsinFig. 3-1 ,wheretheabsorptioncoefcientis0.07mm)]TJ /F6 7.97 Tf 6.59 0 Td[(1.Infact,thisphantompresentsanabsorptioncoefcientequalto0.049mm)]TJ /F6 7.97 Tf 6.58 0 Td[(1inthetargetregion,between0.021mm)]TJ /F6 7.97 Tf 6.59 0 Td[(1and0.07mm)]TJ /F6 7.97 Tf 6.59 0 Td[(1,seventimeshigherthantheabsorptioncoefcientofthebackground.Theseobservationsarequantitativelyconrmedbythevaluecalculatedforthecontrast:3.7215.Observingtheseresults,wecanconcludethatduringthesignalgeneration,theamplitudeoftheinitialpressuredistributionisdeterminedbytheabsorptioncoefcient.TheseresultsconrmthattheamplitudeofthePAsignalisproportionaltotheamountodchromophores. 57

PAGE 58

Figure3-1. Tumormimicphantomexperimentalresult.MIPimageofthephotoacousticsignalsprojectedontheorthogonalplane.Thetargetisthebrightspotbarelyvisibleatthecenteroftheimageanditsabsorptioncoefcientis0.07mm)]TJ /F6 7.97 Tf 6.59 0 Td[(1.Theothertwobrightspotsinthecornersrepresenttworeferencespots,takenasstartingpointtolaunchtheacquisition.Theyarenotpartofthephantom,thushavenottobeconsideredtointerpretetheresults. Figure3-2. Tumormimicphantomexperimentalresult.MIPimageofthephotoacousticsignalsprojectedontheorthogonalplane.Thetargetisthebrightspotbarelyvisibleatthecenteroftheimageanditsabsorptioncoefcientis0.049mm)]TJ /F6 7.97 Tf 6.58 0 Td[(1.Theotherbrightspotinthecornerrepresentsareferencespot,takenasstartingpointtolaunchtheacquisition.Itisnotpartofthephantom,thushasnottobeconsideredtointerpretetheresults. 58

PAGE 59

Figure3-3. Tumormimicphantomexperimentalresult.MIPimageofthephotoacousticsignalsprojectedontheorthogonalplane.Thetargetisthebrightspotbarelyvisibleatthecenteroftheimageanditsabsorptioncoefcientis0.021mm)]TJ /F6 7.97 Tf 6.58 0 Td[(1.Theotherbrightspotinthecornerrepresentsareferencespot,takenasstartingpointtolaunchtheacquisition.Itisnotpartofthephantom,thushasnottobeconsideredtointerpretetheresults. 3.3InVivoPAMExperiments:SubcutaneousVasculatureInthissection,theresultsofthein-vivoPAexperimentstoimagethesubcutaneousvasculatureofthehumanwristareshown.Theprocedureadoptedtoperformtheseexperimentswasdescribedinthesection 2.5 .Foreachsetofexperiments,thesametargetareawasimagedvetimesinarow,eachtimeunderadifferentexternalpressurecondition.Therstacquisitiondidnotinvolvetheuseofthepressuredevice.Intheotherfouracquisitions,describedinsection 2.4 theframe-basedpressuredevicewasusedtoapplyanexternalloadof7000N,15000N,20000Nand25000Nrespectively.Foursetsofexperimentswerecarriedout.Ineachofthefollowingsections 3.3.1 3.3.2 3.3.3 and 3.3.4 weshowtheresultsrelatedtoonesetofexperiment.TheresultsareshownasMaximumIntensityProjectionofthesignalontheorthogonalplane,displayedinacolorscale,rangingfrom0to255.Inthex)]TJ /F1 11.955 Tf 9.3 0 Td[(axisandy)]TJ /F1 11.955 Tf 9.3 0 Td[(axisthenumberofA-lineandB-scanisindicatedrespectively. 59

PAGE 60

Figure3-4. MIPofthestructureofthebloodvesselsintherstsetofexperiments.Thelocatedbloodvesselsareenumeratedfrom1to7.Lettersareusedtoindicatethebranchesofabloodvessel. 3.3.1FirstSetofExperimentsAmapofthestructureofthebloodvesselsinthisareaisshowninFig. 3-4 tosimplifytheunderstandingofthefollowingresults.Thelocatedbloodvesselsareenumeratedfrom1to7.Fig. 3-5 showstheMIPofthephotoacousticsignalgeneratedbythesubcutaneousvasculatureincaseofnoexternalpressureappliedonthepressuredevice.TheMIPsofthephotoacousticsignalgeneratedwhenanexternalloadof7000,15000N,20000Nor25000NwasactingonthepressuredeviceareshowninFig. 3-6 ,Fig. 3-7 ,Fig. 3-8 andFig. 3-9 respectively.Thewaythebloodvesselsrespondtotheactionofthepressurevariesfrombloodvesseltobloodvessel.Forsomeofthem,theamplitudeofthesignaldecreaseswhenthepressureisexertedonthetargetregion.Theresponseofthebloodvessellabeledas1inFig. 3-4 isveryrepresentative.Asitcanbeseenbyobservingthevessel1intheMIPsinFig. 3-5 ,Fig. 3-6 ,Fig. 3-7 andFig. 3-9 ,theamplitudeofthesignaldecreases.Infact,themaximumamplitudeofthesignalinthevessel1awhennopressureis 60

PAGE 61

applied(Fig. 3-6 )is249;ifcalculatedinthesamearea,theamplitudedecreasesto230(Fig. 3-6 )when7000Nareapplied,to210(Fig. 3-7 )when15000Nareappliedandto200(Fig. 3-9 )inthecaseofmaximumpressureexerted.Thus,themaximumamplitudeisreducedby25%whentheloadexertedis25000Nwithrespecttothecasewhennoloadisexerted.Ontheotherhand,forotherbloodvessels,themaximumamplitudeofthephotoacousticsignalincreaseswhenthepressureisexertedonthetargetregion.Thisphenomenonisparticularlyevidentinthesmallerbloodvessellabeledas2.Asitcanbeseenbyobservingthevessel2intheMIPsinFig. 3-6 ,Fig. 3-7 ,Fig. 3-8 andFig. 3-9 ,theamplitudeofthesignalincreases.Infact,themaximumamplitudeincorrespondencetothevessel2whennopressureisapplied(Fig. 3-6 )is60;iftheamplitudeiscalculatedinthesamearea,itincreasesto78(Fig. 3-6 )and75Fig. 3-7 when7000Nand15000Nareappliedrespectively,andto217(Fig. 3-8 )and201(Fig. 3-9 )whenaloadof20000Nor25000Nisappliedrespectively.Thus,themaximumincreaseoftheamplitudeisby261%,measuredcomparingthecaseofnoloadexerted,tothecasewhenaloadof20000Nwasappliedonthepressuredevice.Similarresultscanbeobservedinthebloodvessels5and6inFig. 3-5 .Themaximumamplitudeincorrespondencetobloodvessel5whentheloadis7000N(Fig. 3-6 )is31.7;ifcalculatedinthesamearea,theamplitudeincreasesto52.7(Fig. 3-7 )whentheloadis15000Nandto68(Fig. 3-9 )whentheloadis25000N.Thus,themaximumincreaseoftheamplitudeisby114%,measuredcomparingthecaseof7000Ntothecaseof25000Nappliedonthepressuredevice.Similarly,byanalyzingthemaximumamplitudeincorrespondencetothebloodvessel6,anincreaseby110%canbecomputedcomparingthecasesofexternalloadequalto7000Nand25000N.Furthermore,anotherinterestingphenomenoncanbenoticedobservingthevessels5and6.Thesebloodvesselsarenotevenvisiblewhenthetargetareaisnotsubjecttoanexternalpressure,ascanbeseeninFig. 3-5 ;bycontrast,Fig. 3-6 showsthatsome 61

PAGE 62

partsofthembegintobenoticedwhentheloadis7000N;inFig. 3-9 itcanbeseenthattheirshapebecomesverywell-denedwhenaloadof25000Nexertsapressureonthetargetarea.Otherbloodvesselsshowtrendsimilarto2,5or6,buttheirreactionisslightlydifferent.Infact,inthevessels3and4observedinFig. 3-4 ,theamplitudeofthesignalincorrespondencetothemincreases,butonlywhenaveryhighpressureisexertedonthetargetarea.Thevessel4isveryillustrativeofthisphenomenon.Themaximumamplitudeofthesignalwhennopressureisexerted,is30.8.Ifthisvalueiscomparedwiththeothervaluescalculatedinthesameareainthecasesofloadequalto7000N,15000Nand20000N,itcanbefoundthatthevariationsintheamplitudearelowerthanthe15%.Theamplitudeofthesignalinthisareaincreasesbythe178%whenaloadof25000Nisappliedonthepressuredevice.Infactamaximumamplitudeof85.7ismeasuredincorrespondencetothisbloodvesselunderthemaximumpressureexerted.Otherimportantconsiderationshavetobedoneabouttheramications.Ifwecomparethecasewhennopressureisappliedonthetargetwiththecaseswhenaloadof7000Normoreisexerted,thattheamplitudechangesincorrespondencetoaramication.Particularly,theamplitudeofthesignalincreasesafterthepressureapplication.Thiseffectisevidentintheramicationamongthevesselslabeledas1a,1band1cinFig. 3-4 .Themaximumamplitudemeasuredinthecaseofnoloadactingonthetarget(Fig. 3-5 )is120.Themaximumincreaseisby91%,whichisrelatedtoanamplitudeof230,computedwhenaloadof7000Nisexerted(Fig. 3-6 ).Anothermeaningfulincreaseisfoundwhenaloadof15000N(Fig. 3-7 )or25000N(Fig. 3-9 )isexerted:inthesecasestheamplitudeincreasesbythe75%and66%,sincetheamplitudeis210and200,respectively. 62

PAGE 63

Figure3-5. MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto0N. Figure3-6. MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto7000N. 63

PAGE 64

Figure3-7. MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto15000N. Figure3-8. MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto20000N. 64

PAGE 65

Figure3-9. MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto25000N. Inconclusion,inthissettwotypesofresponsesarefound:(a)largervessels(probablyarterioles)regularlydecreasetheirvolume(andthustheirsignalamplitude)aspressureincreases.Moreover,thedenitionoftheshapeofsomevesselsworsens;(b)conversely,smallvessels(capillaries)increasetheirvolume.Thelatter,sometimesdisplayaregularincreasewhileotheronesdisplayanabruptincreaseaboveathreshold.Moreover,thepressurerevealsthepresenceofothervesselsorimprovesthedenitionofthevesselshape.Similarbehaviorswereobservedinthesecond,third,andfourthsetsofacquisitions,soresultswillbesummarizedfollowingthissubdivisionofvesselswithbehavior(a)andthosewithbehavior(b). 3.3.2SecondSetofExperimentsAsecondsetofmeasureswasperformedinadifferentareaofthewristonthesamesubject.AmapofthestructureofthebloodvesselsinthisareaisshowninFig. 3-10 tosimplifytheunderstandingofthefollowingresults.Thelocatedbloodvesselsareenumeratedfrom1to6. 65

PAGE 66

Figure3-10. MIPofthestructureofthebloodvesselsinthesecondsetofexperiments.Thelocatedbloodvesselsareenumeratedfrom1to6.Lettersareusedtoindicatethebranchesofabloodvessel. Fig. 3-11 showstheMIPofthephotoacousticsignalgeneratedbythesubcutaneousvasculatureincaseofnoexternalpressureappliedonthepressuredevice.TheMIPsofthephotoacousticsignalgeneratedwhenanexternalloadof7000N,15000N,20000Nor25000NwasactingonthepressuredeviceareshowninFig. 3-12 ,Fig. 3-13 ,Fig. 3-14 andFig. 3-15 respectively.Similarlytotherstset,adecreaseoftheamplitudeofthesignalisrecordedforsomebloodvessels(behavior(a)).Thiseffectisevidentbyobservingthetrendoftheamplituderelatedtothebloodvessellabeledas1inFig. 3-10 .Infact,adecreaseoftheamplitudeisrecordedfrom230,calculatedinabsenceofexternalpressure,to185,calculatedincorrespondencetoaloadof20000N.Moreover,theportionofbloodvesselwhereahighamplitudeismeasured,isreducedwhenthepressureisexerted.Inabsenceofexternalpressure(Fig. 3-11 ),asignalwithaveryhighamplitudeismeasuredingreatpartofeachcross-section.Onthecontrary,theportionofeachcross-sectionwhichischaracterizedbyastrongsignalbecomessmallerandsmallerwhenthedifferentpressuresareprogressivelyapplied(Fig. 3-12 toFig. 3-15 ). 66

PAGE 67

Furthermore,becauseofthepressure,theshapeofsomebloodvesselsbecomesirregular.Thishappenstovessels2and4b:theyareverywell-denedinabsenceofpressure,buttheirshapebecomesdiscontinuouswhenaloadhigherthan7000Nisexerted.Underaloadof25000N,thesevesselsarebarelyvisible.Asreportedintherstset,theamplitudeofthesignalincreasesforsomebloodvessels(behavior(b)).Thiseffectissignicantforvessels3cand6:thayarenotevenvisiblewhenthetargetareaisnotsubjecttoanexternalpressure(Fig. 3-11 );whenthepressureincreasestoalevelof20000N,theirshapeisdenedandevenaconnectionbetweenthebloodvessel6andthevessel5emerges.Anincreaseoftheamplitudeisalsorecordedincorrespondencetotheramicationbetweenvesselslabeledas4aand4cinFig. 3-10 .Themaximumincreaseisby214%,whichisrelatedtoanamplitudeof224,computedwhenaloadof7000Nisexerted(Fig. 3-12 ). Figure3-11. MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto0N. 67

PAGE 68

Figure3-12. MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto7000N. Figure3-13. MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto15000N. 68

PAGE 69

Figure3-14. MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto20000N. Figure3-15. MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto25000N. 3.3.3ThirdSetofExperimentsAthirdsetofmeasureswasperformedinadifferentareaofthewristonthesamesubject.AmapofthestructureofthebloodvesselsinthisareaisshowninFig. 3-16 69

PAGE 70

tosimplifytheunderstandingofthefollowingresults.Thelocatedbloodvesselsareenumeratedfrom1to7. Figure3-16. MIPofthestructureofthebloodvesselsinthethirdsetofexperiments.Thelocatedbloodvesselsareenumeratedfrom1to7. Fig. 3-17 showstheMIPofthephotoacousticsignalgeneratedbythesubcutaneousvasculatureincaseofnoexternalpressureappliedonthepressuredevice.TheMIPsofthephotoacousticsignalgeneratedwhenanexternalloadof7000N,15000N,20000Nor25000NwasactingonthepressuredeviceareshowninFig. 3-18 ,Fig. 3-19 ,Fig. 3-20 andFig. 3-21 respectively.Similarlytotherstset,theamplitudeofthesignaldecreasesincorrespondenceofsomebloodvessels(behavior(a)).Forinstance,themaximumamplitudemeasuredinbloodvessel5decreasesby52%whentheloadexertedis20000Nor25000Ncomparedtothecaseinabsenceofpressure.Moreover,incorrespondencetovessel6,asthepressureisapplied,thesignalisnomorerecorded.Thus,thisvesseldoesnotappearinanyoftheMIPsinFig. 3-18 ,Fig. 3-19 ,Fig. 3-20 orFig. 3-21 70

PAGE 71

Asreportedintherstset,thepressureapplicationmakestheamplitudeofthesignalincreaseforsomevesselsorrevealsthepresenceofothervessels(behavior(b)).Incorrespondencetovessel1,inabsenceofexternalpressure(Fig. 3-17 )thecentralportionofeachcross-sectioncharacterizedbyastrongsignalissmaller.Onthecontrary,underanexternalloadof15000N(Fig. 3-19 ),asignalwithaveryhighamplitudeismeasuredingreatpartofeachcross-section.Furthermore,newpartsofvessels1and3becomevisiblewhenaloadof7000Nisappliedonthetargetregion,whereasthesepartsarenotseeninabsenceofpressure.However,assoonashigherpressuresareexerted,theseportionsofvesselsarenomorevisible,andalsosomepartsthatcouldbeobservedinabsenceofpressure,disappear.Similarly,vessel7becomesvisiblewhenaloadof7000Nisapplied(Fig. 3-12 ),whereasonlysmallportions,notevenrecognizableasawholebloodvessel,canbenoticedinabsenceofpressure(Fig. 3-17 ).However,thisvesselisnomorevisibleunderhigherpressuresexerted.Moreover,thebloodvessel4becomesvisibleonlywhenanexternalloadof25000N(Fig. 3-21 )isexerted.Forlowerpressures,thisdoesnotappearintheMIPs. Figure3-17. MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto0N. 71

PAGE 72

Figure3-18. MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto7000N. Figure3-19. MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto15000N. 72

PAGE 73

Figure3-20. MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto20000N. Figure3-21. MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto25000N. 3.3.4FourthSetofExperimentsAfourthsetofmeasureswasperformedinadifferentareaofthewristonthesamesubject.AmapofthestructureofthebloodvesselsinthisareaisshowninFig. 3-22 73

PAGE 74

tosimplifytheunderstandingofthefollowingresults.Thelocatedbloodvesselsareenumeratedfrom1to10. Figure3-22. MIPofthestructureofthebloodvesselsinthefourthsetofexperiments.Thelocatedbloodvesselsareenumeratedfrom1to10. Fig. 3-23 showstheMIPofthephotoacousticsignalgeneratedbythesubcutaneousvasculatureincaseofnoexternalpressureappliedonthepressuredevice.TheMIPsofthephotoacousticsignalgeneratedwhenanexternalloadof7000N,15000N,20000Nor25000NwasactingonthepressuredeviceareshowninFig. 3-24 ,Fig. 3-25 ,Fig. 3-26 andFig. 3-27 respectively.Asreportedalsointheprevioussets,theamplitudeofthesignaldecreasesincorrespondencetosomebloodvessels(behavior(a)).Thisresultcanbeseenobservingthetrendofthebloodvessel2intheupperleftcorneramongthedifferentpressureconditions.Themaximumamplitudemeasuredinabsenceofexternalpressure,whichis123,isprogressivelyreducedassoonasthepressuresareapplied.Themaximumdecreasesisby60%,whenthemaximumloadisexerted(Fig. 3-27 ).Moreover,thepressureapplicationalsoworsenstheshapeofothervessels.Thiseffect 74

PAGE 75

isevidentinvessels4,5,6andespecially9,thatisnomorevisiblewhenaloadof7000Nisexerted.Ontheotherhand,anincreaseofthevesselvolumeisreportedforothervessels(behavior(b)).Thebloodvessellabeledas1inFig. 3-23 representsanexample.Inabsenceofpressure(Fig. 3-23 ),themaximumamplitudemeasuredcharacterizestheinferiorhalfofthevolumewhichconstitutesthevessel.Whenthedifferentpressuresareexerted,thevolumeassociatedwithsuchhighamplitudeincreasesprogressively:underanexternalloadof20000Nand25000N(Fig. 3-26 andFig. 3-27 ),theamplitudeofthesignalisthehighestalongthewholevessel.Asfurthereffectofthepressureapplication,newportionsofvesselsbecomevisibleonlywhentheloadisapplied:thisphenomenoncanbeseenbyacomparisonofthevessel10inabsenceofexternalpressure(Fig. 3-23 )withthecasewhenwhentheloadis20000N(Fig. 3-26 );underthispressureevensomeconnectionsappearbetweenwithvessels8and9.Moreover,thepressureapplicationenhancestheshapeofsomevesselsorbranchpoints:forinstancethebranchpointbetweenvessel2and3isverywell-denedunderaloadof20000Nor25000N(Fig. 3-26 andFig. 3-27 respectively),whereastheshapeofthispartissketchedintheMIPobtainedinabsenceofpressure(Fig. 3-23 ). 75

PAGE 76

Figure3-23. MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto0N. Figure3-24. MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto7000N. 76

PAGE 77

Figure3-25. MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto15000N. Figure3-26. MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto20000N. 77

PAGE 78

Figure3-27. MIPimageofthephotoacousticsignalfromthesubcutaneouswristvasculatueprojectedontheorthogonalplaneincaseofexternalloadequalto25000N. 78

PAGE 79

CHAPTER4DISCUSSIONInthischapter,wecommenttheresultsofthein-vivoPAMimagingofthesubcutaneousvasculatureunderdifferentexternalpressures,showninsection 3.3 .First,weexplainthehaemodynamicchangesintermsofowrate.Afterwards,suchchangesareinterpretedbyexplainingseveralpossiblephenomenathatmighthavegeneratedthesehaemodynamicresponsesafterthepressureapplication.Thepressureapplicationdoesnotgeneratethesameeffectamongthebloodvesselsintheimagingarea.Thewaythevesselsreactvariesfrombloodvesseltobloodvessel.Thoseresultscanbedividedintotwobiggroups,accordingtothetrendoftheamplitudeofthesignal.Ononehand,ithasbeenfoundthat,forsomebloodvessels,theamplitudeofthesignaldecreaseswhenthepressureisexertedonthetargetregion.Moreover,itcanbeseenthatothereffectsareassociatedwiththisresult:whenthemaximumamplitudeofthesignaldecreases,theportionofeachcross-sectionwhichischaracterizedbyastrongsignalbecomessmallerandsmallerwhenthedifferentpressuresareprogressivelyapplied.Alsotheportionofbloodvesselwhereahighamplitudeismeasured,isreducedwhenthepressureisexerted.Furthermore,somebloodvesselsarevisibleintheMIPsonlyinabsenceofpressure.Assoonasthepressureisexerted,thesevesselsdisappearfromtheMIPs.Alternatively,onlyportionsofthemremainvisible,buttheyarenotevenrecognizableasawholebloodvessel.Whattheseresultshaveincommon,isthedecreaseoftheamplitudeofthesignalintheMIPs.Suchamplitudeisproportionaltotheamplitudeofthephotoacousticsignal.Adecreaseofthisamplitudecanbeascribedtoareductionofthenumberofchromophoresthatabsorbthelightandgeneratethephotoacousticsignal,asexplainedinthetheoryinsection 1.1.1.1 andvalidatedbythephantomexperimentsinsection 2.3.1 .Sincehaemoglobinanddeoxyhaemoglobinareconsideredasthemajorchromophoresinthebloodand 79

PAGE 80

sinceanexcitationlightwithanisosbesticwavelengthhavebeenused,changesintheamplitudeofthePAsignalresultfromchangesinthetotalhaemoglobinconcentration.Thus,adecreaseintheamplitudeofthesignalisduetoalocaldecreaseofthequantityofbothspeciesofhaemoglobin,whichmeansadecreaseofthebloodvolume.Basingontheseresultsandaccordingtotheaforementionedphotoacoustictheory,wecaninferthatforasignicantamountofbloodvessels,theapplicationoftheexternalpressureprovokesadecreaseofthebloodvolume.Iftheseresultsarereadintermsofimagingquality,adecreaseoftheamplitudeofthesignalimpliesadeteriorationoftheimagequality.First,areductionoftheamplitudeofthesignalfromthebloodvessels,determinesaworseningofthecontrastc,whichisdenedasc=AROI)]TJ /F4 7.97 Tf 6.58 0 Td[(Abackground AbackgroundwhereAROIisthearithmeticmeanofthenormalizedamplitudesofthesignalincorrespondencetotheregionofinterest(ROI),whichistheregionwherethetargetislocated;Abackgroundisthearithmeticmeanofthenormalizedamplitudesofthesignalincorrespondencetothebackground.Becauseofthiseffect,afterthepressureapplication,edgesandcontoursofbloodvesselsappearlesssharpandsometimesthevesselsarenotwelldistinguishablefromthebackground.Alsoanothereffectdeterminesaworseninginthenalimage.Afterthepressureapplication,thePAsignalfromcertainbloodvesselsisnomoredetected.Thus,thesevesselsarenotshownintheMIPoronlyportionsofthemremainvisible.Therefore,theanatomicalinformationfromthosevesselsislost.Thedecreaseofthebloodvolumecanbedescribedbytheuidodynamictheory[ 58 59 ].ThevolumetricowrateQinacylindricalductisdescribedbythePoiseuille'slaw:Q=P R (4) 80

PAGE 81

wherePisthepressuredifference,orpressuredrop,betweenthetwoendsandRistheresistencetoow.RcanbeexpressedasinEqn. 4 :R=8l r4 (4)whererandlaretheradiusandthelengthoftheductrespectively,istheviscosityiftheliquidowinginsidetheduct.TheEqn. 4 canbeobtainedbywritingEqn. 4 explicitly:Q=Pr4 8l (4)AccordingtothePoiseuille'slaw,thevolumetricowrateisproportionaltotheradiusandtothepressuredifference,andisinverselyproportionaltotheuidviscosityandthelengthoftheduct.Theresistancedependslinearlyupontheviscosityandthelength,butthefourthpowerdependenceupontheradiusisdramaticallydifferent.Thustheradiusoftheductgivesthemostimportantcontributiontothevolumetricowrate.TheseconsiderationsleadtoexpectadecreasedvolumeofvesselsintheROI,whenapressureisexertedbytheperipheralringonavesselsegmentwithinwardowtotheROI.TheresultingpressuredropoverthecompressedsegmentwouldinduceareductionintransmuralpressureoftheimagedsegmentwithintheROIand,accordinglyreducedradiusandvolume.Flowwouldalsobereducedinthiscase,butnocontrastisassociatedtothisparameterwithPAimaging.ItshouldberemarkedthatPoiseuille'slawallowsthecalculationoftheuidynamicresistanceonlyunderpreciseconditions(newtonian,continuousandlaminarow)whicharenotrespectedparticularlyinsmallvessels(bloodisanon-newtonianuid,becauseitiscomposedofredbloodcells,whitebloodcellsandplateletssuspendedinbloodplasma)andalsothatthearterialowisnotcontinuousbutpulsatileandlaminarowcanbedisruptedbytheringinducedstenosis.Stillthehypothesisofthepressure 81

PAGE 82

ringinducingstenosisandapressuredropinthesubsequentvesselsegmentcanbeassumedqualitativelyvalid.Ontheotherhand,whereasforsomebloodvesselsthepressureapplicationmakestheamplitudeofthesignaldecrease,theoppositeeffectoccursforthemajorityofthebloodvessels.Forthesevessels,theamplitudeofthesignalincreaseswhenthepressureisexertedonthetargetregion.Moreover,otherphenomenaareassociatedwiththisresult.Inthevesselswhereanincreaseoftheamplitudeofthesignalisfound,theportionofeachcross-sectionwhichischaracterizedbyastrongsignalbecomeslargerwhenthedifferentpressuresareprogressivelyapplied.Incertaincases,alsoalongerportionofvesselcanbeassociatedwithahighersignallevelwhenthepressureisexerted.Furthermore,somebloodvesselsthatdonotappearintheMIPinabsenceofpressure,becomethenvisible.Finallyalsoanincreaseoftheamplitudeofthesignalincorrespondencetoabranchpointisobservedafterthepressureapplication.AlltheseresultshaveincommontheincreaseoftheamplitudeofthesignalintheMIPs.Asalreadyexplainedabove,suchamplitudeisproportionaltotheamplitudeofthephotoacousticsignalandultimatelytobloodvolumewithnodistinctionbetweenoxygensaturationlevelofblood,sinceanisosbesticwavelengthwasuse.Basingontheseresultsandaccordingtotheaforementionedphotoacoustictheory,wecaninferthatforasignicantamountofbloodvessels,theapplicationoftheexternalpressureprovokesanincreaseofthevesselvolume.Iftheseresultsarereadintermsofimagingquality,manyimageenhancementscanbeobservedincorrespondencetothosebloodvesselswhereanhigheramplitudeisrecorded.First,thosevesselsareshownintheMIPwithastrongercontrast,becausetheAROI,thearithmeticmeanofthenormalizedamplitudesofthesignalintheregionofinterest,increases.Therefore,theshapeofthebloodvesselsisshownintheMIPswithhigherdenition.Moreover,theedgesofthevesselsappearmoresharp.Otherimportantadvantageouseffectshavebeenreported.Somevessels,whosesignalis 82

PAGE 83

notrecordedinabsenceofpressure,becomevisibleintheMIPonlyafterthepressureapplication.Forinstance,thepressureapplicationinsomecasesrevealsconnectionsbetweentwobloodvesselsorhighlightsthepresenceofbranchpoints.Similarly,theproleofmanyvesselscanbeoutlinedonlywhenthetargetareaissubjecttotheexternalpressure;otherwise,onlysomesmallpartscanbeseenintheMIPandthevesselstructureishardtoassess.Alltheseimprovementsconsidered,westatethatthepressureapplicationgivesanimportantcontributiontotheimagingofthebloodvesselsstructure,becauseitallowstheidenticationsofimportantvascularfeatures.Sucharesultbecomesmorerelevantifweconsiderthatmanydifferentfeaturesofvascularitypermitthedistinctionbetweenmalignantandbenignprocesses.Asexplainedinsection 1.2.2 incaseofmalignanttumor,tumorvasculatureisdisorganizedandhastrifurcationsandbrancheswithunevendiameters;thetumorvesselwallexhibitsstructuralalterations;manyarterio-venousshuntingmakethevasculatureabnormalandchaotic;highvasculartortuosityandvasodilatationoccur;bloodperfusionintumorsisspatiallyandtemporallyheterogeneous.Manyphenomenahavebeeninvestigatedtoexplainthenatureofthehaemodinamicresponsesgeneratedbythepressureapplication.Fromourresearch,itemergesthatnosinglefactorisbyitselfasufcientexplanationfortheincreasedbloodvolume.Atrivialhypothesiscouldbethatoftheringexertingpressure(hencecausingstenosis)onsegmentswithoutwardowfromtheROI,thusinducinganincreasedtransmuralpressureandvolumeintheimagedsegment.However,thispurelyhemodynamichypothesisseemslargelyinsufcienttoexplainthegeneralizesignalincreaseinsmallvesselsandmechanismsrelatedtolocalautoregulationeffectsshouldbeconsideredasapossiblecausesifnotthemainones.First,theincreaseofowratemightbeduetoalocalpressure-inducedcutaneousvasodilation.Thisisaslowrespondingandtransientadaptivephenomenon,initiatedbyawiderangeofpressure 83

PAGE 84

changes.Responsestolocalmechanicalstressesaremediatedthroughaconsiderablenumberofcutaneousreceptors.Someofthesereceptorsareconnectedtothinbersthat,onceactivated,canproducealocalvasodilationresponseintheskin.AccordingtoFromyetal.,localpressurestraintotheskinmightplayanimportantroleincutaneousmicrocirculatoryimpairment[ 60 ].Indeed,thisphenomenonisthoughttobearesponsetolocalexternalpressureapplication,bywhichtheskinisprotectedfromischemiaduringmechanicalstimulation[ 61 ].AccordingtoAbrahametal.,theinduced-pressurevasodilationisinitiatedbyanon-noxiouspressurestimulation[ 61 ].Intheirwork,thecutaneousbloodowinthehumanhandwasmeasuredbyusinglaserDopplerowmetryandanexternalramppressurewasappliedthroughtheprobetip.Theyfoundthatalocalandprogressivenon-noxiouspressurestimulationresultsinareexincreaseofbloodowintheskinofthehumanhand.Fromyetal.supportedtheseresultsandprovedthatthishyperemiaismediatedbytheactivationoftheunmyelinatedslowCbers.Infact,ithasbeenfoundthatCberreceptortypesintheprimateskinexhibitparticularresponsestopressuresrangingfrom0.7g/mm2to2.5g/mm2,duringprogressiveexternallyappliedpressurestrain[ 62 ].Fromyetal.showedthatthesemechanoreceptorsnerveterminalsnotonlyrespondtonon-nociceptivestimulibutalsoparticipatetonoxiousandnociceptivevasodilation,producedbystimuliabovetheringthresholdlevelofnociceptors[ 62 ].Inagreementwiththesendings,Whiteandal.foundthatsustainedstimuliwhichareinitiallyabovethepainthreshold,mayinduceaconstantlow-levelactivityinCbersnociceptors,afteraconsiderabledelayofuptominutes.Eveninresponsetoasustainedsuprathresholdmechanicalstimulus,Cbersexhibitaninitialhigh-frequencydischargefollowedbyaconstantlowlevelactivity.Inbothcases,thepressure-inducedvasodilationresultsalong-lastingphenomenon.AlsoAbrahametal.foundthatthevasodilatoreffectofpressurestimulationcontinuesseveralminutesafterthestimulusisremoved[ 61 ]. 84

PAGE 85

Ontheotherhand,vasodilationcouldalsoresultfromaresponsetotissue-derivedmetabolicfactorsratherthanamechanoreceptoractivation.Theveryrsteffectpressureapplicationmightbeareductionoftheowrate,whichultimatelycausesareductionoftheoxygenintake.Sincethetissuestilltakesandconsumestheoxygentransportedbythevascularsystem,adecitofoxygenoccurs.HaddyandScottfoundthatfactorsrelatedtoskeletalmusclecelldepolarizationinitiatethehyperemia,whilefactorsrelatedtooxygenconsumptioninexcessofoxygendeliverycontributemoretothemaintenanceofthehyperemia[ 66 ].BlissandToth,inagreementwiththeseresults,foundthatreactivehyperemiaiscloselyrelatedtooxygendebt[ 67 68 ].AccordingtoToth'sexperiments,theincreaseinreactivehyperemiawithocclusionlastingminutesisduetooxygenlackandaccumulationofmetabolicwasteproducts[ 67 ].Tothetal.alsofoundthatvasodilatorproductsofanaerobicmetabolismarethoughttoincludeadenosine,lacticacidandH+.AfallinpHincreasestheconductanceofthevascularsmoothmuscleATP-sensitiveK+channel;elevatedlactatehasbeenshowntocausearteriolarvasodilationbytheactiononcGMP;adenosineactsonspecicreceptorsofvascularsmoothmuscleandmaycausedilationbyincreasingtheconductanceofATP-sensitiveK+channels,increasingsynthesisofnitricoxide(NO)andprostaglandisbytheendothelium.Interestingly,NOisthoughttohaveanimportantroleinthereactivehyperemia.Blissfoundthatthemicrocirculationiscontrolledbyvasoactivesubstancessecretedlocallybytheendothelialcells.ThemostimportantisNOwhichfacilitatesowbycausingrelaxationofvascularsmoothmuscle[ 68 ].Tagawaetal.examinedtheroleofNOinthecontrolofvasculartoneduringreactivehyperemiainhumanforearmvessels.TheyfoundthatNOisapotentvasodilatoraswellasastrongendothelium-derivedrelaxingfactoranditisinvolvedinthelatephaseofreactivehyperemia.Infact,ithasbeenfoundthattheincreaseinowcausestheincreaseinshearstressthatreleasesNOfromtheendothelium[ 69 ]. 85

PAGE 86

Asalreadymentioned,theresultsinourin-vivoexperimentsareheterogeneous.Forcertainvessels,adecreaseofthebloodowwasreported,whileothervesselsexhibitedtheoppositetrend.Asemergesfromourliteraturestudy,manycomplexphysiologicalphenomenamightbeinvolvedwiththebloodvolumereductionaswellaswithavasodilation.Besides,eachphenomenonmightberegulatedbyagreatvarietyofpathways.Itisdifculttounambiguouslyexplainwhybloodvesselsinthesameareashowsuchadifferentresponse.Probablythepressuredeviceexertsamechanicalstressonthebloodvessels,byreducingtheirlumenandreducingthetransmuralpressurebetweentwoendofavessel.Suchchangesimplyaowratereductionforasignicantnumberofbloodvessels.Themechanicalstressduetothepressureapplicationmighttriggerapressure-inducedvasodilationforagroupofbloodvessels.Thedilatationeffectduetoanon-nociceptivemechanismisthemostlikelyone.However,althoughoursubjectdidnotreportanyunpleasentsensation,avasodilatatoryeffectduetoanociceptiveactivationisnotdiscardedinourstudy.Thelong-lastingfeatureofthepressure-inducedvasodilation,wouldexplainwhythiseffectisobservedinourresultsevenafterseveralacquisitionsonthesamearea.Anyway,aconsequenceofthisvasodilationmightbetheincreaseoftheshearstressalongthebloodvessel'swall,whichimpliesareleaseofNOfromtheendothelium.TheNOisthoughttobeinvolvedinthemid-latephaseofthereactivehyperemia,thusitislikelytoberesponsibleforsuchalonglastingvasodilationfoundinourin-vivoexperiments.Furthermore,anothereffectofthepressureapplication,mightbeareductionoftheowrate,whichcausesareductionoftheoxygenintakeandultimatelyanoxygendebit.Thistissuehypoxiamightinducehyperemiaasacompensatorymechanismtoprovidemoreoxygentothetissueandtoremovewasteproducts.Theseassumptionsaresupportedbytheaforementionedliteraturestudies.However,furtherstudiesareneededtofullyexplainthephenomenaobservedinourstudy. 86

PAGE 87

Animportantcommenthastobedoneabouttheplanningandthedurationofourexperiments.Asalreadyexplainedinsection 2.5 ,foreachsetofexperiments,5acquisitionsinarowwereperformed.Eachofthemlasted15minutes.Therstacquisitionwasdonetoimagethevasculatureinabsenceofexternalpressure.Ineachoftheotherfouracquisitions,aloadof7000N,15000N,20000Nand25000Nrespectivelywasappliedtothestructure.Eachacquisitionwasstartedrightaftertheendofthepreviousone,thusnotimewaslefttothevasculaturetoreturntoitsinitialconditioninabsenceofpressure.Undertheseconditions,asetofexperimentslasted90minutes.Wecannotexcludethattheconsecutiveacquisitionsandthedurationofthewholeexperimentdidnotalterorinuencetheactionofthepressuredeviceonthetissue.Abrahametal.measuredthecutaneousbloodowinthehumanhandbyusinglaserdopplerowmetryandanexternalramppressurewasappliedthroughtheprobetip,foramaximumof60minutes[ 61 ].Theirstudysupportsthepressure-inducevasodilationafternon-noxiouslocalpressurestimulation.Noothersimilarresults,supportingeitherthiseffectorthemetabolicresponse,arefoundintheliterature. 87

PAGE 88

CHAPTER5CONCLUSIONANDFUTUREDEVELOPMENTSImagingthevasculatureprovidesapowerfultoolforstudyingthevasculaturedevelopment.Thiskindofimagingbecomesmuchmoreimportantifitisaddressedtothestudyoftumorangiogenesis,becausesuchimagingprovidesusefulinformationaboutthetumordevelopmentaswellasthenature,benignormalignant,ofthetumor.Photoacousticmicroscopywaschosentoachievethisgoal,becausethistechniqueisnon-invasiveandprovidesgreattissuedifferentiation.Toobtainastructuralimageofthevasculature,thetotalhaemoglobinconcentrationwasconsidered.Thus,anisosbesticwavelengthwaschosen,atwhichhaemoglobinanddeoxyhaemoglobinexhibitthesameabsorptioncoefcient.PAMwasusedtoin-vivoimagethestructureofthesubcutaneousvasculatureinthehumanwristandtoevaluatehowsuchkindofimagingismodiedbytheapplicationofanexternalpressureonthetarget.Wemanufacturedatailoreddevicethatwasincontacttothewristsurfaceandexertedamechanicalpressureonthetissue.ByusingPAM,wedetectedlocalvariationsofowrateandbloodvolumegeneratedbytheapplicationoftheexternalpressureandevaluatedthedifferencesbetweenthenalimagesobtainedwithoutandwithpressure.Accordingtoourndings,theresponseofbloodvesselstothepressureapplicationvariesfromvesseltovesselandcouldbedividedintotwogroups.Ononehand,thepressureapplicationgeneratedadecreaseoftheamplitudeofthePAsignal,whichrevealsalocalreductionofthebloodvolume.Incorrespondencetothebloodvesselswherethiseffectisreported,adeteriorationoftheimagequalitycanbeobserved.Infact,thecontrastdecreasesandsomevesselsorpartsofthemarenomoreimaged.Inthiscase,weconcludedthatthepressureapplicationdidnotpositivelycontributetotheimagingofthesubcutaneousvasculature.Ontheotherhand,theoppositeeffectoccurredforthemajorityofthebloodvessels.Inthesecases,thepressureapplicationgeneratedanincreaseofbloodvolume 88

PAGE 89

andowrate,revealedbyanincreaseoftheamplitudeofthePAsignal.Thiseffectallowedabetterimagingofthevasculature,becausetheshapeofthevesselscanbeobservedwithahigherdenitionandtheedgesappearsharper.Moreover,thepressureapplicationinsomecasesrevealedthepresenceofbloodvesselsthatwerenotimagedinabsenceofpressure.Inthiscase,wecanconcludethatthepressureactivationgaveanimportantcontributiontotheimagingofthesubcutaneousvasculature,becausecrucialanatomicalfeaturesofthevasculatureareenhancedinthenalimage.AfuturedevelopmentmayconsistinusingPAMtoestimatetherelativeconcentrationofhaemoglobin(Hb),anddeoxyhaemoglobin(HbO2),inthebloodvesselsinabsenceofexternalpressureandwhenthepressureisapplied,andcomparetheresults.ThisfunctionalapplicationofPAMispossible,becauseHbandHbO2havedifferentspectra.Sincetheconcentrationoftwomoleculeshastobeestimated,accordingthespectroscopytheory,anexcitationlightwithatleasttwowavelengthshastobeusedtoseparatethecontributionfromHbandHbO2ThisapplicationofPAMimagingallowsimportantfunctionalstudiesaboutthebloodoxygenationandsaturationaswellastheoxygenconsumption.TheestimationoftherelativeconcentrationofHbandHbO2,aswellasthecomputationofthebloodsaturationhaveinterestingapplications.Infact,suchatechniquecouldbeusedtoassesstheoxygenationofthevasculaturegrownaroundatumor,becausehypoxiaisoneofthefunctionalhallmarkstodistinguishbenigntumorsfrommalignanttumors.Furthermore,thisfunctionalapplicationofPAMcouldbeusedtoassessandverifythepresenceoflocalhypoxiathatisthoughttogenerateafterthepressureapplicationinourinvivoexperiments,asdiscussedinchapter 4 .Anotherimportantfuturedevelopmentinvolvesanalternativeorganizationofourexperiments.Asexplainedinsection 4 ,inthecurrentstudy,eachacquisitionwasstartedrightaftertheendofthepreviousoneandnotimewaslefttothevasculature 89

PAGE 90

toreturntoitsinitialconditioninabsenceofpressure.Wecannotexcludethattheconsecutiveacquisitionsandthedurationofthewholeexperimentdidnotalterorinuencetheactionofthepressuredeviceonthetissue.Inafuturestudy,beforestartingeachacquisition,enoughtimecouldbelefttothecirculationtoreturntoitsrestcondition.Evenifthiswouldprolongthedurationofthewholesetofexperiment,acomparisonbetweentheresultsandthecurrentresultswouldbeinterestingtoevaluateanypossibledifference. 90

PAGE 91

REFERENCES [1] A.G.Bell,OntheProductionandReproductionofSpeechbyLight,AmericanJournalofScience,vol.20,No.118,pp.305,Revised1880. [2] J.Tyndall,ActionofanIntermittentBeamofRadiantHeatuponGaseousMatter,ProceedingsoftheRoyalSocietyofLondon,vol.31,pp.307,1881. [3] Roentgen,Ontonesproducedbytheintermittentirradiationofagas,PhilosophyMagazine,Series5,vol.11,pp.306,1881. [4] A.Rosencwaig,Theoryofthephotoacousticeffectwithsolids,inJournalofAppliedPhysics,vol.47,1976. [5] A.G.Bell,UpontheProductionofSoundbyRadiantEnergy,PhilosophyMagazine,Series5,vol.11,pp.510,1881. [6] W.H.Preece,Ontheconversionofradiantenergyintosonorousvibrations,ProceedingsoftheRoyalSocietyofLondon,vol.31,pp.506,1881. [7] M.E.Mercadier,Surlaradiophonieproduiteal'aideduselenium.ComptesRendusdel'Acadl'miedesSciences,,vol.92,pp.705,1881. [8] P.Beard,Biomedicalphotoacousticimaging,Interfacefocus,pp.602,2011. [9] L.V.Wang,Multiscalephotoacousticmicroscopyandcomputedtomography,Naturephotonics,vol.3,pp.503,2009. [10] L.V.Wang,TutorialonPhotoacousticMicroscopyandComputedTomography,IEEEJournalofSelectedTopicsinQuantumElectronics,vol.14,pp.171,2008. [11] L.H.V.Wang,X.Zhao,H.SunandG.Ku,Microwave-inducedacousticimagingofbiologicaltissues,inReviewofScienticInstruments,vol.70,pp.3744,1999. [12] C.HaischandR.NiessnerLightandSoung:photoacousticspectroscopy,inSpectroscopyEurope,vol.14,2012. [13] rsfs.royalsocietypublishing.org [14] X.Xu,H.Liu,H.SunandL.H.V.WangTime-riversedultrasonicallyencodedopticalfocusingintoscatteringmedia,inNaturePhotonics,vol.16,2011. [15] S.SakadzicandL.H.V.Wang,CorrelationTransferandDiffusionofUltrasound-ModulatedMultiplyScatteredLight,inTheAmericanPhysicalSociety,vol.96,2006. [16] L.H.V.Wang,Prospectofphotoacoustictomography,inMedicalPhysics,vol.35,2008. 91

PAGE 92

[17] L.H.V.WangandS.HuPhotoacoustictomography:invivoimagingfromorganellestoorgans,inScience,vol.335,2012. [18] A.FensterandD.B.Downey,-Dultrasoundimaging:areview,inIEEEEngineeringinMedicineandBiologyMagazine,vol.15,pp.41,1996. [19] W.Denk,J.H.StricklerandW.W.Webb,Two-photonlaserscanninguorescencemicroscopy,Science,vol.248,73,1990. [20] D.Huang,E.A.Swanson,C.P.Lin,J.S.Schuman,W.G.Stinson,W.Chang,M.RHee,T.Flotte,K.Gregory,C.A.PullafttoandJ.GFujimoto,OpticalCoherencetomography,Science,vol.254,pp.1178,1991. [21] H.F.Zhang,K.Maslov,G.StoicaandL.V.andWang,Functionalphotoacousticmicroscopyforhigh-resolutionandnoninvasiveinvivoimaging,Naturebiotech-nology,Vol24,pp.848,2006. [22] K.Maslov,G.StoicaandL.V.andWang,Invivodark-eldreectionmodephotoacousticmicroscopy,OpticalSocietyofAmerica,Vol30,N.6,2005. [23] H.F.Zhang,K.MaslovandL.V.andWang,Invivoimagingofsubcutaneousstructuresusingfunctionalphotoacousticmicroscopy,Natureprotocols,Vol2,pp.797,2007. [24] C.Kim,K.HSong,F.GaoandL.V.Wang,SentinelLymphNodesandLymphaticVessels:NoninvasiveDual-ModalityinVivoMappingbyUsingIndocyanineGreeninRats-VolumetricSpectroscopicPhotoacousticImagingandPlanarFluorescenceImaging,Radiology,vol.255,pp.442,2010. [25] X.Cai,W.Li,C.Kim,Y.Yuan,L.V.WangandY.Xia,Invivoquantitativeevaluationofthetransportkineticsofgoldnanocagesinalymphaticsystembynoninvasivephotoacoustictomography,inACSNano,vol.5,pp.9658,2011. [26] E.W.Stein,K.MaslovandL.V.andWang,Noninvasivemappingoftheelectricallystimulatedmousebrainusingphotoacousticmicroscopy,6856,68561JJ,2008. [27] LaserInstituteofAmerica,AmericanNationalStandardforSafeUseofLasersANSIZ136.1-2000,(AmericanNationalStandardsInstituteInc.,NewYork,NY,2000). [28] L.D.Liao,M.L.Li,H.Y.Lai,Y.Y.I.Shih,Y.C.Lo.S.TsangandP.C.PChao,Imagingbrainhemodynamicchangesduringratforepawelectricalstimulationusingfunctionalphotoacousticmicroscopy,inNeuroimage,vol.52,,pp.562,2010. 92

PAGE 93

[29] H.F.Zhang,K.Maslov,L.V.andWang,Automaticalgorithmforskinproledetectioninphotoacousticmicroscopy,inJournalofbiomedicaloptics,vol.14,2009. [30] H.F.Zhang,K.Maslov,M.Sivaramakrishnan,L.V.andWang,Imagingofhemoglobinoxygensaturationvariationsinsinglevesselsinvivousingphotoacousticmicroscopy,inAppliedPhysicsLetters,vol.90,2007. [31] X.Lei,Multi-scalephotoacoustictomographyanditsapplicationstocancerresearch,2012. [32] A.A.Karabutov,N.B.Podymova,andV.S.Letokhov,Time-resolvedlaseroptoacoustictomographyofinhomogeneousmedia,inAppliedPhysicsB:LasersandOptics,vol.63,pp.545,1996. [33] A.A.Karabutov,E.V.Savateeva,andA.A.Oraevsky,Optoacoustictomography:Newmodalityoflaserdiagnosticsystems,inLaserPhysics,vol.13,pp.711-AS723,2003. [34] G.Ku,X.D.Wang,X.Y.Xie,G.Stoica,andL.H.V.WangImagingoftumorangiogenesisinratbrainsinvivobyphotoacoustictomography,inAppliedOptics,vol.13,pp.711,2003. [35] X.Lei,Lightscattering,2012 [36] F.HillenandA.W.Grifoen,Tumourvascularization:sproutingangiogenesisandbeyond,inCancerMetastasisRev,vol.44,pp.770,2005. [37] S.M.WaisandD.A.Cheresh,Tumorangiogenesis:molecularpathwaysandtherapeutictargets,inNatureMedicine,vol.17,N.11,2011. [38] F.NussembaumandI.M.Herman,Tumorangiogenesis:InsightsandInnovations,inJournalofOncology,vol.2010,2009. [39] J.Folkman,TumorAngiogenesis:TherapeuticImplications,inTheNewEnglandJournalofMedicine,vol.285,pp.1182,1971. [40] J.Folkman,TumorAngiogenesis,inAdvancesincancerresearch,vol.19,pp.331,1974. [41] D.D.ZerbinoandM.DmitrukIntratumoralproliferationofthebloodvessels,inArkhPatologii,vol.45,pp.80,1983. [42] S.Oladipupo,S.Hu,A.C.Santeford,J.Yao,J.Kovalski,R.V.Shohet,K.Maslov,L.V.WangandJ.M.ArbeitConditionalHIF-1inductionproducesmultistageneovascularizationwithstage-specicsensitivitytoVEGFRinhibitorsandmyeloidcellindependence,inBlood,vol.117,pp.13264,2011. 93

PAGE 94

[43] J.Laufer,P.Johnson,E.Zhang,B.Treeby,B.Cox,B.PedleyandP.Beard,Invivopreclinicalphotoacousticimagingoftumorvasculaturedevelopmentandtherapy,inJournalofBiomedicalOptics,vol.17,pp.4142,2012. [44] H.M.Jensen,I.Chen,M.R.DeVaultandA.E.Lewis,Angiogenesisinducedbynormalhumanbreasttissue:aprobablemarkerforprecancer,'inScience,vol.218,pp.293,1982. [45] Q.Zhu,M.Huang,N.Chen,K.Zarfos,B.Jagjivan,M.Kane,P.HedgeandS.H.Kurtzman,Ultrasound-guidedopticaltomographicimagingofmalignantandbenignbreastlesions:initialclinicalresultsof19cases,inNeoplasia,vol.5,2003. [46] D.FukumuraandR.K.Jain,Imagingangiogenesisandthemicroenvironment,inAPMIS,vol.116,pp.695,2008. [47] D.E.Feldman,V.ChauhanandA.C.Koong,Theunfoldedproteinresponse:anovelcomponentofthehypoxicstressresponseintumors,inMolCancerRes,vol.3,pp.597,2005. [48] J.Yao,K.I.MaslovandL.V.Wang,Invivophotoacoustictomographyoftotalbloodowandpotentialimagingofcancerangiogenesisandhypermetabolism,inTechnologyinCancerResearchandTreatment,vol.11,pp.301,2012. [49] J.Laufer,P.Johnson,E.Zhang,B.Treeby,B.Cox,B.PedleyandP.Beard,Invivopreclinicalphotoacousticimagingoftumorvasculaturedevelopmentandtherapy,inJournalofBiomedicalOptics,vol.17,2012. [50] H.Anderson,P.Price,M.Blomley,M.O.LeachandP.Workman,Measuringchangesinhumantumourvasculatureinresponsetotherapyusingfunctionalimagingtechniques,inJournalofBiomedicalOptics,vol.17,2012. [51] http://openi.nlm.nih.gov/detailedresult.php?img=3005160_boe-1-1-310-g002&req=4 [52] K.Maslov,H.F.Zhang,S.HuandL.V.Wang,Optical-resolutionphotoacousticmicroscopyforinvivoimagingofsinglecapillaries,inOpticsLetters,vol.33,pp.929,2008. [53] H.F.Zhang,K.Maslov,M.L.Li,G.StoicaandL.V.Wang,Invivovolumetricimagingofsubcutaneousmicrovasculaturebyphotoacousticmicroscopy,inOpticsExpress,vol.14,pp.9317,2006. [54] D.Bergstrom,TheAbsorptionofLaserLightbyRoughMetalSurfaces,2008. [55] http://www.mathworks.com/ [56] http://www.vsg3d.com/amira/overview 94

PAGE 95

[57] http://www.itksnap.org/pmwiki/pmwiki.php [58] S.M.Richardson,FluidMechanics,inHemisphere,NewYork,1989. [59] J.HappelandH.Noordhoff,LowReynoldsNumberHydrodynamics,Noordhoff,Leyden,1973. [60] B.Fromy,P.AbrahamandJ.L.Saumet,Non-nociceptivecapsaicin-sensitivenerveterminalstimulationallowsforanoriginalvasodilatatoryreexinthehumanskin,inBrainResearch,vol.811,pp.166,1988. [61] P.Abraham,B.Fromy,S.Merzeau,A.JardelandJ.L.Saumet,Dynamicsoflocalpressure-inducedcutaneousvasodilationinthehumanhand,inMicrovascularResearch,vol.61,pp.122,2001. [62] B.Fromy,P.AbrahamandJ.L.Saumet,Non-nociceptivecapsaicin-sensitivenerveterminalstimulationallowsforanoriginalvasodilatoryreexinthehumanskin,inBrainResearch,vol.811(1-2),pp.166,1998. [63] M.D.Gee,P.LynnandB.Cotsell,TherelationshipbetweencutaneousCbretypeanantidromicvasodilatationintherabbitandrat,inJournalofPhysiology,vol.503,pp.31,1997. [64] R.Schmidt,M.Schmeltz,H.E.TorebjorkandH.O.Handwerker,Mechano-insensitivenociceptorsencodepainevokedbytonicpressuretohunamskin,inNeuroscience,vol.98,pp.793,2000. [65] D.M.WhiteandJ.D.Levine,DifferentmechanicaltransductionmechanismsfortheimmediateanddelayedresponsesodratC-bernociceptors,inJournalofNeurophysiology,vol.66,pp.363,1991. [66] F.J.HaddyandJ.D.Scott,Metabolicfactorsinperipheralcirculatoryregulation,inJournalofNeurophysiology,vol.34,1975. [67] A,Toth,M.Pal,M.IntagliettaandP.C.Johnson,Contributionofanaerobicmetabolismtoreactivehyperemiainskeletalmuscle,inAmericanJournalofPhysiology-HearthandCirculatoryPhysiology,vol.292,2006. [68] M.R.Bliss,Hyperaemia,inJournaloftissueviability,vol.8,1988. [69] T.Tagawa,T.Imaizumi,T.Endo,M.Shiramoto,Y.HarasawaandA.Takeshita,RoleofNitricOxideinReactiveHyperemiainHumanForearmVessels,inCirculation,1994. 95

PAGE 96

BIOGRAPHICALSKETCH MariaCristinaLetiziawasborninMessina,Italyin1989.SheobtainedherBachelorofScienceinbiomedicalengineeringin2011fromPolitecnicodiMilano,Italy.ShecontinuedherpostgraduatestudyatPolitecnicodiMilanofocusingonherworkonelectronictechnologiesinbiomedicalengineering.AsarecipientoftheAtlantisCRISPdualdegreegrantshersttookclassesatPolitecnicodiMilano,Italy,andthenmovedtotheUniversityofFloridatojointheJ.CraytonPruittFamilyDepartmentofBiomedicalEngineering,Florida,USAtocompleteherMasterofSciencedegreeinbiomedicalengineeringunderthesupervisionofDr.HuabeiJiang.UponcompetitionofherstudiesatboththePolitecnicodiMilanoandtheUniversityofFloridasheplanstopursueacareerintheresearcheld. 96