Accuracy ofventricular volume estimation

SUMMARY Sources oferrorintheestimation ofventricular volumefromCAT scans are discussed andtheaccuracyofthemeasurementassessed. Twomethods ofcalculating ventricular volumefroman intraventricular injection ofradioisotope aredescribed andcompared. Finally, theCAT scan andisotope methods arecompared andcontrasted. Theerrorassociated witha single measurementofvolumebyany ofthese techniques isbetween 20%and30%.Inpatients withno ventricular catheter there isno choice other thantouse CAT scanning, butifintraventricular injection ispossible thismethodoffers a potentially more accuratevolume determination because ofhishigher signal-to-noise ratio. Computerised tomographic (CAT)scanning provides anewmethodofassessing ventricular size. Becausethistechnique isatraumatic (unless anaesthesia isrequired) itislikely tobeusedmuch morefreely thanprevious methods, andinparticular repeated examinations inthesamepatient will befeasible. Thismakesitimportant toknow thereliability ofassessments ofventricular size based onCAT scans, andinparticular toknowthe significance ofchanges insize between different examinations onthesamepatient (Gawler etal., 1976; Roberts etal., 1976; Pennetal., 1978). Estimates ofventricular sizeareimportant in diagnosis, indeciding abouttheneedtointervene surgically (inparticular byventricular shunting procedures), andinjudging theefficacy oftreatmentdesigned todealwithhydrocephalus. The rangeofchanges inventricular size encountered in clinical practice hasledmostpeople tofeelthat decisions canbetakenwithout anexact measure ofventricular size. Whilethis maybeso,there are likely tobeanincreasing numberofcircumstances inwhichprecise measurements will beofvalue. In normalpressure hydrocephalus ithasbeensuggested thattheresults ofshunting donotcorrelate withareduction inventricular size (Shenkin etal., 1975; Jacobs andKinkel, 1976), buttheaccuracy of theestimation ofsize mustbecarefully considered instudies ofthisnaturebeforedrawing such conclusions.