ABSTRACT
ORIGINAL AR
TICLE
INTRODUCTION Defecography is a radiological evalu-ation of the anatomy and function of the anorectum and pelvic floor. Initially describedbyWalldénin1952,itisrecog-nizedasavaluablediagnosticmethodfor theevaluationofanorectalphysiologyand evacuationdisorders.1-3
Techniques for the examination vary significantlybetweenmedicalinstitutions, althoughmostphysiciansattempttofollow standardsrecommendedbyMahieuetal.4 Rectaloutliningwithathickbariumpaste isfollowedbyfluoroscopy,andradiographs are usually acquired to register anorectal eventsduringthreephases:squeeze,strain andrest.Duringthisexamination,several measurements (anorectal angle, anorectal junction,puborectalismusclelength,anal canallengthanddegreeofanalrelaxation) areusuallyobtainedthroughdirectexami-nationoftheX-rayfilms.5,6
Atourinstitution,themethodologyof videotapingthefluoroscopicimagesisusu-ally employed (videodefecography). After digitizing the image, computer software called ANGDIST is used for determining themeasurementparameters.Hence,X-ray filmisseldomobtained,andthetechniqueis knownascomputerizedvideodefecography.
OBJECTIVE This study was undertaken to compare measurementsobtainedbydirectobservation of radiographs and through computerized image analysis (computerized videodefe-cography), in a prospective manner using independentobservers.
METHOD Thepresentstudyunderwentevaluation andwasapprovedbytheethicscommittees
oftheDepartmentofGastroenterology(Uni-versidadedeSãoPaulo)andouruniversity hospital (Hospital das Clínicas, Faculdade deMedicinadaUniversidadedeSãoPaulo). Full informed consent was obtained from allparticipants.
Ten asymptomatic individuals volun-teered to undergo videodefecography. Six individuals were female.Their ages ranged from20to47years,withameanof25(stan-darddeviation=2.9years).Theseindividuals hadnohistoryofpreviouspelvicoranorectal surgery,withoutcomplaintsofconstipationor incontinence,andtheysaidthattheywereable tomovetheirbowelseverydaywithoutthe aidoflaxatives.Theydidnothaveanyclini-callydetectedanorectaldiseaseandpresented normalsigmoidoscopy.
All individuals underwent standard videodefecography examination performed bythesameradiologyteam.Theequipment wastheCGR(CompagnieGénéraledeRadi- ologie)TILTIX,connectedtoavideomoni-torandaprofessionalVHShi-fivideotape recorder(IntermedVideoTechnologiesInc., modelGXR-70UHI-FI).Thecomposition anddensityofthebariumpasteutilizedwere thesameforallindividuals.Thebariumwas injecteduntilthepatientssaidthattheyhad thesensationofafullrectum(avolumeof approximately200ml).2
Thefollowingmeasurementsweretaken: • anorectalangle(ARA):anglebetweenthe
axesoftheanalcanalandrectum(taken atthebackoftherectalwall).Thiswas calculatedduringthethreephasesofthe examination:squeeze(ARA-s),rest(ARA-r)anddefecation(ARA-d).
• anorectal junction (ARJ): distance be-tweentheanorectaljunctionandaline drawnfromthetipofthecoccyxtothe inferior border of the pubic bone.This wasalsoobtainedduringthethreephases (ARJ-s,ARJ-randARJ-d).
Computerizedvideodefecography
versusdefecography:
doweneedradiographs?
FaculdadedeMedicinadaUniversidadedeSãoPaulo,SãoPaulo,Brazil
CarlosWalterSobrado CarlosEduardoFonsecaPires SergioEduardoAlonsoAraújo EdsonAmaro
AngelitaHabr-Gama DesidérioRobertoKiss
SaoPauloMedJ.2005;123(3):105-7.
CONTEXTANDOBJECTIVE:Defecographyhas been recognized as a valuable method for evaluating patients with evacuation disorders. Itconsistsoftheuseofstaticradiographyand fluoroscopytorecorddifferentsituationswithin anorectaldynamics.Conventionally,rectalpa-rametersaremeasuredusingradiograms.Itis rareforfluoroscopyalonetobeused.Computer softwarehasbeendevelopedwiththespecific aim of calculating these measurements from digitized videotaped images obtained during fluoroscopy,withouttheneedforradiographic film,therebydevelopingacomputerizedvideo-defecographymethod.Theobjectivewasthus to compare measurements obtained via com-puterizedvideodefecographywithconventional measurementsandtodiscusstheadvantagesof thenewmethod.
DESIGN AND SETTING:Prospective study at theradiologyserviceofHospitaldasClínicas, UniversidadedeSãoPaulo.
METHOD: Ten consecutive normal subjects underwentvideodefecography.Theanorectal angle,anorectaljunction,puborectalismuscle length,analcanallengthanddegreeofanal relaxation were obtained via the conven-tional method (using radiography film) and via computerized videodefecography using the ANGDIST software. Measurement and analysis of these parameters was performed bytwoindependentphysicians.
RESULTS:Statisticalanalysisconfirmedthatthe measurements obtained through direct radiog-raphyfilmassessmentandusingdigitalimage analysis (computerized videodefecography) wereequivalent.
CONCLUSIONS: Computerizedvideodefecogra-phyisequivalenttothetraditionaldefecography examination. It has the advantage of offering reducedradiationexposurethroughsavingon theuseofradiography.
106
• C-line (C)/puborectalis muscle length: linedrawnfromtheinferiorborderofthe pubicbonetothepointontheposterior rectal wall with greatest flexion due to puborectalis muscle contraction. C was obtained during the three phases (C-s, C-r,andC-d).
• analcanallength(ACL) • analcanalrelaxation(ACR)
Foreachindividualexamined,twoinde-pendentobserverscalculatedthesevariables: oneobserverusingconventionalradiographic proctograms (defecography) and the other usingcomputerizedimageanalysis.
Duringcomputerizedvideodefecography analysis, the ANGDIST computer software wasusedfordeterminingthesemeasurements from the digitized videotape images.This software is able to calculate distances and anglesfromanyimage(intheformofbitmap files),withpropercalibrationofthedistances and pincushion distortion adjustment.The softwareiscalibratedbymeasuringtheimage ofametalsphereof1.9cmindiameterthat ispositionedabovethepatient’spubisduring theexamination.
TheANGDISTsoftwarewascreatedby theengineerPauloEduardoPilonusingC++ programminglanguage,anditrunsunderthe Windows® operatingsystem.Furtherinforma-tionaboutthissoftwaremaybeobtainedby gettingintouchwithitscreator,viathee-mail address[email protected].
For statistical analysis, variance analysis (ANOVA) was used to compare differences between the mean values of measurements obtainedviathetwomethods.TheSpearman rankcorrelation(rs)wasusedtodeterminethe
degreeofcorrelationbetweenmeasurements obtained via the two methods.The signifi-canceleveladoptedwas5%(p=0.05).
RESULTS In this series, there was no significant difference between the mean values of the measurementsobtainedviaconventionalde-fecographyandimageanalysis(computerized videodefecography),aspresentedinTable1.
For all measurements, the Spearman correlationindexwasclosetoone,asshown inTable1.
DISCUSSION Defecography has increasingly been recommendedfortheevaluationofanorectal functiondisorders,especiallyforconstipation and incontinence.6,7 It is a noninvasive and welltolerateddiagnostictoolthatiscapable ofprovidingvaluableinformationregarding patients’pelvicandanorectalphysiology.5
Several measurements can be made via radiological examination, in particular the anorectal angle, anal canal length, degree of anal relaxation, puborectalis muscle length,perinealdescentanddegreeofrectal evacuation.Thesemeasurementsareusually obtained during defecography by means of direct calculations using radiographs taken followingfluoroscopicscanning,sincevisual assessmentofthefluoroscopyvideomonitor for this purpose is frequently hampered by imagedistortion.2,6
Withtheaimofovercomingthisdifficulty anddoingawaywiththeneedforacquisition ofradiographs,theANGDISTcomputersoft-warewasdeveloped.Thisprogramprovides
Table1. Meanvaluesandcorrelationsofmeasurementsobtainedduringconven-tionaldefecography(CD)andcomputerizedvideodefecography(CVD)
Parameters meanvalue correlation
CD* CDV† p r† p
ARA-s 75.5 76 0.947 0.84 0.012
ARJ-s 2.27 2.13 0.715 0.97 0.004
C-s 5.37 5.49 0.796 0.91 0.006
ARA-r 114.35 114.9 0.931 0.93 0.005
ARJ-r 3.7 3.85 0.822 0.92 0.006
C-r 7.47 7.76 0.662 0.99 0.003
ARA-d 139.35 134.3 0.521 0.98 0.003
ARJ-d 4.67 4.66 0.985 0.88 0.009
C-d 8.7 8.93 0.700 0.99 0.003
ACL 1.87 1.91 0.765 0.84 0.012
ACR 1.19 1.25 0.623 0.92 0.006
*ARAvaluesindegrees.Allothermeasurementsincentimeters. †Spearmanrankcorrelation.
ARA=anorectalangle;ARJ=anorectaljunction;C=puborectaliscontraction;s=squeeze;r=rest;d=defecation;ACL= analcanallength;ACR=analcanalrelaxation.
angle and distance calculations during the analysisofimagesobtainedbydigitizingana-logfluoroscopicimagesrecordedonvideotape. Since ANGDIST uses independent vertical andhorizontalcalibration,imagedistortion isautomaticallycorrected.2
The present study has demonstrated strongcorrelationbetweenthemeasurements (ARA, ARJ and C during the three phases of the examination, and ACL and ACR) obtainedviaconventionalanalysis(defecog-raphy) and image analysis (computerized videodefecography).Thetwotypesofanalysis wereperformedbydifferentobservers,with oneblindedtotheresultsoftheother.
Thedegreeofrectalevacuationwasnot evaluatedinthepresentstudy,sincetheuse ofgraphicalcomputeranalysisforobtaining thisresulthadalreadybegunatourinstitu-tionin1998.2,8
InspiteofthereportpublishedbyMahieu etal.4in1984,regardingvideotaperecording offluoroscopicimagesobtainedduringdefe-cography(videodefecography),whichshowed thatsuchatechniqueenablesexcellentdynam-icevaluationofanorectalandpelvicfunction, the analysis of radiographic images remains the preferred method for the calculation of defecographic variables. Nevertheless, the acquisitionofradiographsincreasespatients’ exposure to radiation beyond the exposure neededforfluoroscopyrecording.
Sincefluoroscopicimagescanberecorded and stored, the development of a method thatusestheseimagesforthecalculationof parameters,withnoneedforadditionalX-ray exposure,washighlydesirable.9-11Afterstudy ofthefeasibilityofdigitalcaptureofvideo-defecography images, it became possible to devisecomputersoftwarethatwascapableof definingmeasurementviaimageanalysis.
Mostindividualsundergoingdefecogra-phyareyoungpeopleatreproductiveageand, althoughtheradiationdoseduringdefecogra-phylargelyremainswithinacceptablelevels, computerized videodefecography does away withadditionalX-rayexposure.3,11Moreover, the additional costs involved in performing videodefecography no longer represent a significantburden,sincecomputerhardware isbecomingmoreaccessible.
CONCLUSIONS We conclude that videodefecography is feasible, provides defecographic parameters equivalent to conventional defecography and does away with the need for acquiring radiographs and the consequent additional exposuretoradiation.
107
1. Walldén L. Defecation block in cases of deep rectogenital pouch; a surgical roentgenological and embryological study withspecialreferencetomorphologicalconditions..ActaChir Scand.1952;165:1-122.
2. Sobrado Junior CW. Contribuição da videodefecografia di-nâmicacomputadorizadanoestudodedoentessubmetidosà graciloplastia.[thesis]SãoPaulo(SP):FaculdadedeMedicina daUniversidadedeSãoPaulo;1999.
3. GoeiR.Defecography:principlesoftechniqueandinterpreta-tion.Radiologe.1993;33(6):356-60.
4. MahieuP,PringotJ,BodartP.Defecography:I.Descriptionof anewprocedureandresultsinnormalpatients.Gastrointest Radiol.1984;9(3):247-51.
5. WiersmaTG,MulderCJ,ReedersJW,TytgatGN,VanWaes PF.Dynamicrectalexamination(defecography).BaillieresClin Gastroenterol.1994;8(4)729-41.
6. JorgeJM,Habr-GamaA,WexnerSD.Clinicalapplicationsand techniquesofcinedefecography.AmJSurg.2001;182(1):93-101. 7.
CuriLA,MauriziM.Valordiagnósticodeladefecografia.[Di-agnosticvalueofdefecography].ActaGastroenterolLatinoamer. 2001;31(4):313-7.
8. KarlbomU,NilssonS,PahlmanL,GrafW.Defecographicstudy ofrectalevacuationinconstipatedpatientsandcontrolsubjects. Radiology.1999;210(1):103-8.
9. GoeiR,KemerinkG.Radiationdoseindefecography.Radiol-ogy.1990;176(1):137-9.
10. ParryRA,GlazeSA,ArcherBR.TheAAPM/RSNAphysicstuto-rialforresidents.Typicalpatientradiationdosesindiagnostic radiology.Radiographics.1999;19(5):1289-302. 11. ZoncaG,DeThomatisA,MarchesiniR,etal.Doseassorbita
dalle gonadi dei pazienti adulti sottoposti a studio defeco-grafico con acquisizione radiografica digitale o tradizionale [Theabsorbeddosetothegonadsinadultpatientsundergoing defecographicstudybydigitalortraditionalradiographicimag-ing].RadiolMed(Torino).1997;94(5):520-3.
Sourcesoffunding:None Conflictofinterest:Notdeclared Dateoffirstsubmission:May20,2003 Lastreceived:February2,2005 Accepted:February9,2005
AUTHOR INFORMATION
CarlosWalterSobrado,MD,MSc,PhD. Professor,Colorec-tal Surgery Division, Department of Gastroenterology, FaculdadedeMedicinadaUniversidadedeSãoPaulo, SãoPaulo,Brazil.
CarlosEduardoFonsecaPires,MD.DivisionofAlimentary Tract Surgery, Department of Gastroenterology, Facul-dadedeMedicinadaUniversidadedeSãoPaulo,São Paulo,Brazil.
Sergio Eduardo Alonso Araújo, MD, MSc.Division of Ali-mentaryTractSurgery,DepartmentofGastroenterology, FaculdadedeMedicinadaUniversidadedeSãoPaulo, SãoPaulo,Brazil.
Edson Amaro, MD, PhD.Department of Radiology, Facul-dadedeMedicinadaUniversidadedeSãoPaulo,São Paulo,Brazil.
Angelita Habr-Gama, MD, PhD.Professor, Division of Ali-mentaryTractSurgery,DepartmentofGastroenterology, FaculdadedeMedicinadaUniversidadedeSãoPaulo, SãoPaulo,Brazil.
Desidério Roberto Kiss, MD, PhD.Professor and Head of DivisionofAlimentaryTractSurgery,DepartmentofGas-troenterology,FaculdadedeMedicinadaUniversidadede SãoPaulo,SãoPaulo,Brazil.
Addressforcorrespondence: CarlosWalterSobrado
R.Itapeva,500—Conjunto7-B—BelaVista SãoPaulo(SP)—Brasil—CEP01332-000 Tel.(+5511)3253-1616/3252-0249—Fax(+55 11)3253-0174
E-mail:[email protected] E-mail:[email protected]
Copyright©2005,AssociaçãoPaulistadeMedicina
REFERENCES
SaoPauloMedJ.2005;123(3):105-7.
RESUMO
Videodefecografiacomputadorizadaversusdefecografia:asradiografiassãonecessárias?
CONTEXTOEOBJETIVO: O exame de defecografia tem sido reconhecido como método valioso para avaliaçãodepacientescomdistúrbiosdaevacuação.Adefecografiaconsisteemregistrar,pormeiode fluoroscopiaeradiografiasestáticas,diferentessituaçõesdadinâmicaanorretal.Nométodoconvencio-nal,asradiografiassãoutilizadasparaocálculodeparâmetrosretais.Éraraautilizaçãoapenasde fluoroscopia.Umprogramadecomputadorfoidesenvolvidoparacalcularessesparâmetrosatravésda digitalizaçãodasimagensregistradasemvídeopelafluoroscopia,criandoummétododevideodefeco-grafiacomputadorizada.Assim,oobjetivofoidecompararosvaloresdeparâmetrosdadefecografia calculadospelométodocomputadorizadopropostocomaquelesobtidospormétodoconvencionalede discutirasvantagensdonovométodo.
TIPODEESTUDOELOCAL:Estudo prospectivo no serviço de Radiologia do Hospital das Clínicas da UniversidadedeSãoPaulo.
MÉTODO:Dezindivíduosvoluntáriosnormaisforamsubmetidosaoexamededefecografia,noqualforam obtidas,apartirdométodoconvencional(pormeioderadiografias)edométodocomputadorizado(vide-odefecografiacomputadorizada),asseguintesmedidas:ânguloanorretal,junçãoanorretal,comprimento domúsculopuborretal,comprimentodocanalanalegraudeaberturadoânusemcadaumadasfases doexame.Asavaliaçõeseaanálisedosparâmetrosdefecográficosacimadescritosforamrealizadas pordoisobservadoresmédicosindependentes.
RESULTADOS:Osresultadosobtidos,apósanáliseestatística,comprovaramaequivalênciadométododa videodefecografiacomputadorizadacomparadoaométodoconvencional.
CONCLUSÕES: Avideodefecografiacomputadorizadaémétodoequivalenteaométodotradicionaldede-fecografiaquepermitemenorexposiçãodopacienteàirradiaçãopordispensarousoderadiografias
PALAVRAS-CHAVE:Defecografia.Fisiologia.Ânus.Radiação.Cirurgiacolorectal.