Page 128 - Demo

126Chapter7VIRTUALGUIDANCEUsingapenciltheuserwasinstructedtomarkthe(virtual)planningpointsthatwerevisualisedbytheHoloLensonthe3DmodelsAQRmarkerwasattachedtothepencilsotheHoloLenscouldalsotrackthepencil’sposition ThisallowedtheHoloLenstodisplayavirtualarrowshowingtheuserhowthepencilshouldbemanipulatedtotargetthetipofthepenciltotheplanningpoint Theredcolourofthevirtualarrowturnedorangeiftheoffsetwasbelow3mmandgreenifitwasbelow2mmInadditiontheabsolutedistancebetweenthepenciltipandreferencepointwasindicatedontheuser’stop-leftview(Figure3)ACCURACYSTUDYTwoobserverswhowereskilledHoloLensusersperformedtwoexperimentstwiceInthefirstexperimenttheobserversplacedthesurgicalguidesonthe10cranialshapesanddemarcated10pointswithapencilAdigitalcalliperwith001mmaccuracywasusedtomeasurethe2Ddistancebetweenthemarkedpointsand(atleast)5beaconpointsUsing thesedistancesand the trilaterationprinciplethe3Dpositionwasreconstructed19 TheerrorreportedinthisstudywasdefinedastheEuclideandistancebetweenthemarkedandplannedpoints Theobserversrepeatedtheexperimenttwoweekslatertoexamineintra-observervariabilityInthesecondexperimentthesameobserversusedtheHoloLensapplicationtotransfertheplanAftercalibrationandregistration10holographicpointsweredemarcatedwithapencilonthesurfaceofthe3D-printedcranialshapesErrorswerecalculatedusingthebeaconpointssimilartotheapproachinthefirstexperimentandtheexperimentwaslikewiserepeatedtwoweekslaterbytheobserversStudent’s t-tests withasignificancelevelofalpha=005 wereused tocompare thesizesof theerrorsBackwardslinearregression wasused todetermine the variablesthatmostinfluencedtheoutcome Theintraclasscorrelationcoefficient(ICC)wascalculatedtodeterminetheconsistencyofthemeasuringtechnique