Eur J Dent Educ. 2019;00:1–9. wileyonlinelibrary.com/journal/eje  | 1© 2019 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd

Received:17August2018  |  Accepted:23May2019DOI: 10.1111/eje.12447

O R I G I N A L A R T I C L E

Exploring training dental implant placement using computer‐guided implant navigation system for predoctoral students: A pilot study

Janina Golob Deeb1 | Sompop Bencharit2,3  | Caroline K. Carrico4 | Marija Lukic5 | Daniel Hawkins6 | Ksenija Rener‐Sitar5,7 | George R. Deeb6

1DepartmentofPeriodontics,SchoolofDentistry,VirginiaCommonwealthUniversity,Richmond,Virginia,USA2DepartmentofGeneralPractice,SchoolofDentistry,VirginiaCommonwealthUniversity,Richmond,Virginia,USA3DepartmentofBiomedicalEngineering,SchoolofEngineering,VirginiaCommonwealthUniversity,Richmond,Virginia,USA4OralHealthPromotionandCommunityOutreach,OralHealthResearchCore,VirginiaCommonwealthUniversity,Richmond,Virginia,USA5DivisionforDentalMedicine,MedicalFaculty,UniversityofLjubljana,Ljubljana,Slovenia6SchoolofDentistry,DepartmentofOralandMaxillofacialSurgery,VirginiaCommonwealthUniversity,Richmond,Virginia,USA7DepartmentofProsthodontics,UniversityDentalClinics,UniversityMedicalCentreofLjubljana,Ljubljana,Slovenia

CorrespondenceSompopBencharit,DepartmentofGeneralPractice,SchoolofDentistry,VirginiaCommonwealthUniversity,Richmond,VA23298‐0566.Email:sbencharit@vcu.edu

AbstractIntroduction: Recentcomputer‐guideddynamicnavigationsystemspromiseanoveltrainingapproachforimplantsurgery.Thisstudyaimedtoexaminelearningprogressinplacementofdentalimplantsamongdentalstudentsusingdynamicnavigationona simulation model.Materials and Methods: Seniorstudentswithnoimplantplacementexperiencewererandomlyassignedfiveimplantplacementattemptsinvolvingeitherthreemaxillaryorfourmandibularimplantsdistributedintheanterior/posterior,andleft/rightseg‐ments.ImplantplacementwasplannedusingaNavidentDynamicGuidancesystem.Surgical timewas recorded. Horizontal, vertical and angulation discrepancies be‐tweentheplannedandplacedimplantpositionsweremeasuredusingsuperimposedCBCTscans.Datawereanalysedwith repeatedmeasures regressionwithTukey'sadjustedpairwisecomparisons(α = 0.05).Results: Fourteenstudentsparticipated,withameanageof26.1yearsandequalmales and females. Mean time for implant placement was associated with at‐temptnumber(P <0.001),implantsite(P = 0.010) and marginally related to gender (P =0.061).Studentshadasignificantreductionintimefromtheirfirstattempttotheirsecond(10.6vs7.6minutes;adjustedP <0.001)thenplateaued.Overall3Dan‐gulation(P <0.001)and2Dverticalapexdeviation(P =0.014)improvedwitheachat‐tempt,butchangesinlateral2D(P =0.513)andoverall3Dapexdeviations(P = 0.784) werenotstatisticallysignificant. Implantsiteswereassociatedwith lateral2D,2Dverticalandoverall3Dapexdeviation(P < 0.001).Discussion: Malesweremarginallyfasterthanfemales,hadslightlyloweroverall3Dangulation,andreportedhigherproficiencywithvideogames.Noviceoperatorsim‐provedsignificantlyinspeedandangulationdeviationwithinthefirstthreeattemptsofplacingimplantsusingdynamicnavigation.Conclusion: Computer‐aideddynamic implantnavigation systemscan improve im‐plantsurgicaltraininginnovicepopulation.

K E Y W O R D S

dentalimplants,education,navigation,simulation,students,surgery

2  |     GOLOB DEEB Et aL.

1  | INTRODUC TION

Implantpositioninginrelationtoplanneddefinitiveprosthesescanbeenhancedusingcomputer‐guidedstaticordynamicsystems.1‐5 Staticguidedimplantplacementsurgeryinvolvestheuseofaconebeam computed tomography (CBCT) generated surgical guidewithmetalsurgical tubes.Thesestaticguidescaneitherbesup‐portedbyadjacentnaturalteeth,mucosaoralveolarbone.1‐5Staticguided surgery has been shown to be more accurate than freehand implant placement.1‐5 Recent development of inexpensivethree‐dimensional (3D)printersallowscost‐effectivestaticguidefabricationandthereforehavepopularisedthemethod.4 Implantpositioningispredeterminedinastaticguide;however,thestaticguide does not allow for real‐time adjustmentswhen needed orvisualisation of the osteotomy.While tooth‐supported ormuco‐sal‐supportedstaticguidedsurgeryisindicatedwithflaplesssur‐gerywhenbonegraftingorosseousmodification is notneeded,staticguidedsurgerycanbedifficultinpatientswithlimitedmouthopening,implantsiteswithdifficultaccessordirectvisualisation,aswellasimplantplacementinlimitedhorizontalspacesbetweenadjacentteeth.5

Dynamicnavigationsurgeryallowstheoperatortofullyvisualisetheosteotomyandimplantsiteonthecomputerscreenwhilepre‐paringtheosteotomysiteandplacinganimplantfixture.Theaccu‐racyofdynamicnavigationhasbeenobservedtobecomparabletothatofstaticguidedplacement.6Dynamicallyguidedimplantplace‐menthasbeenshown tobemoreaccurate than freehand implantplacement in terms of angular deviation, platformpositioning andapicalpositioning.1,7‐9Mostdynamicguidedimplantsurgerystudieshoweverhavebeenperformedbyexperiencedsurgeonswithpriortraining on the respective navigation system.1,6 The questionwasraised whether dynamic navigation technology could be used totrainthenoviceoperator,suchasadentalstudentwithnopreviousimplantsurgicalexperience,toperformimplantplacementcompe‐tently and accurately.

This prospective randomised studywasdesigned to evaluatethe learning progression, defined as accuracy in placement ofdental implants on a simulationmodel,when a computer‐guideddynamicnavigationwasusedtotrainseniordentalstudentswithnopreviousimplantplacementtraining.Computer‐guideddynamicnavigation, such as Navident Dynamic Navigation used in thisstudy,utilisesvirtualsimulationandprovidesimmediatefeedbackduringtheimplantplacementthroughinteractiveCBCT3Dmod‐elling. The studywas designed to define the learning curve andtheminimalnumberofattemptsnecessary inutilisingcomputer‐guided implant navigation system to improve implant placementskill in a novice implant trainee. Themain hypothesiswas therewas a statistical difference in implant sitesmeasuredby the im‐provementofimplantplacementaccuracyandimplantplacementtime.Additionally,thesecondaryhypothesiswasthattherewasastatistical difference betweenmales and females based on theirimplantplacement improvementmeasuredby the implantplace‐ment accuracy and time.

2  | MATERIAL AND METHODS

The study protocolwas approvedby the university InstitutionalReviewBoard,IRBNo.HM20011878.Seniordentalstudentswithnopriorsurgicalimplantplacementexperiencewererecruited.Atotal of five implant placement attemptswere assigned to eachstudent.Thefirstfourattemptswererandomlyassignedforplac‐ing either three maxillary or four mandibular implants. Implantsites included maxillary first right molar, maxillary right centralincisor,maxillary left firstmolar andmandibular left firstmolar,mandibularleftsecondpremolar,mandibularrightsecondpremo‐lar,mandibular right firstmolar. The implant planningwas doneby the consultation of three authors who were board‐certifiedprosthodontist (Author‐SB), board‐certified periodontist (AuthorJGD) and board‐certified oral andmaxillofacial surgeon (AuthorGRD).Theimplantsitedistributionswerealsorandomlyassignedforanterior/posterioraswellas left/rightonasimulationmodel.Onthefifthattempt, thestudentsplaced implantsonboth jawsrandomlyassignedpositions (anterior/posteriorand left/right).Arandomisation schedulewasgenerated inSASEGv6.1 software(SASInstitute)toassignallstudentstoarandomsequenceofthefirstfourjaws(twoformaxillaandtwoformandible),aswellastheimplantsiteswithineachjawforallattempts.Therandomisationfirstassignedthestudentstoaparticularjawandthenrandomisedthe implant siteswithin that jaw to reduce the logistical burdenthatrandomisingtheimplantsiteswouldhavebrought(ie,chang‐ing jawsbackandforth).Usingtherandomisationschedule,ninestudentsbeganwith themandibleand five studentsbeganwiththemaxilla.Aftereachattempt,theparticipantscompletedques‐tionnairesregardingto(a)previousdentalsimulationexperience,simulations for caries preparation or restorative indications; (b)priorvideogamingexperience;and (c)perceiveddifficultywhenusingthenavigationsystemtoassistimplantplacementinregardstojaws(maxilla/mandible),sides(left/right)andimplantsites(an‐terior/posterior). Prior to the first attempt, an orientation wasdoneforeachparticipanttoexplainhowthestudywouldbedone,how thenavigation systemworked andhow touse implant sur‐gical handpieceanddrills.Aone‐weekwashoutperiodbetweeneachattemptwasused.

Polymethylmethacrylate 3D printed maxillary and mandibu‐larmodelswere taggedwith three fiducialmarkings,oneplacedon buccal aspect apically to the central incisors and the othertwoplacedontheposterioraspectofthemodel.Thesemarkersensured the three‐dimensional orientation of the model in theCBCT and enabling accurate superimposition. A pre‐operativeCBCT scan using iCAT FLX V10 (Imaging Sciences InternationalLLC)of themodelwas takenwitha radiographicstentand fidu‐cialmarkers inplace.Thedata fromCBCTwere loaded into theNavident(Claronav)dynamicguidancesystemsoftwarewherevir‐tual implantplacement includingappropriate implant size (widthand length) aswell as implant positionwas planned. The spatialmatching of the model to its virtual on‐screen representation(CBCTimage)wasregistered.Thespatialrelationshipbetweenthe

     |  3GOLOB DEEB Et aL.

JagTagonthestentinstalledonaplasticmodelandtheDrillTagonhandpiecewastrackedbythestereoscopiccamera.(Figure1)Theregistrationallowedforthecontinuoustrackingofthejawduringthe navigated osteotomy and formaintaining its accuracy if thejawmoves. (Figure2)NotethatthestudyusedcommonmanikinusedintheUSdentalschools.Allcasesweredonewiththerub‐bersimulatedcheeks/softtissue.However,itwasdifficulttotakeaphotographofdrillangulationwiththerubbercheeks inplace.Theimplantplacementfigurethereforehadnorubbercheekstobetterdemonstratethedrillangulationandpositioning.Thedigitalprostheticsetupswereperformedandthecrownpositionswereusedforimplantplanning.

Themodelandthestentwiththepatternedjawtagweretightlysecuredontothemannequin.TheNavidentdevicewasplacedinfrontoftheoperatorwiththecameraabovetheoperatingfield.Appropriate positioning of the mannequin, camera, computerscreenwasensuredtoenableeasyvisualisationofthecomputerscreendisplayingreal‐timefeedbackofthedrillinrelationtotheplannedimplantposition.Trackingsystemarrayfromthecameratothejawtagduringtheosteotomypreparationwasusedtoaccu‐ratelylocatethepositionofthehandpieceinrelationtothemodelandscan.(Figure1)Priortoosteotomypreparationforeachdrill,drillcalibrationwasperformedonthejawtagtoprovidethesys‐temappropriatedrilllength.Thisinturnsprovidedinformationondepthof theosteotomypreparation.A real‐timevideo feedbackthroughoutsimulation,inrelationtotheplannedimplantposition(withappropriatedesireddepthandangulation),wasusedtoguidethe implantsitepreparationandplacement.Theaccuracyof thedrillpositionandangulationinrelationtotheplannedpositionoftheimplantwasmonitoredandthedeviationswereallertedusingadifferentialcolourcodedsystem(Figure2).

Operating timewasrecordedforeachattempt.AllmodelswerescannedinCBCTfollowingimplantplacement.Superimpositionofthepre‐operativescanwithplannedimplantpositionandthepost‐opera‐tivescanwithplacedimplantpositionwasperformedusingEvaluNav(Claronav) software through the three fiduciary markers (Figure 3).Theplannedandplacedimplantpositionswerecompared(Figure4).Repeated measures ANOVA was used to evaluate discrepancies inthetwo‐dimensional(2D)lateraldeviation,overall3Dapexdeviation,2Dverticalapexdeviationandoverall3Dangledeviation (Figures4and5).Allmodelsadjustedfortheattemptnumber(1‐5),implantsite(maxillaryrightfirstmolar,maxillaryrightcentralincisor,maxillaryleftfirstmolar,mandibularleftfirstmolar,mandibularleftsecondpremo‐lar,mandibularrightsecondpremolarandmandibularrightfirstmolar)andthegenderoftheoperator.PosthocpairwisecomparisonswereadjustedforusingTukey'sHSD.SASEGv.6.1(SASInstitute)wasusedforallanalyses.SignificancelevelwassetatP =0.05.Thenullhypoth‐esiswas that there isno statisticaldifference inoutcomemeasures(time,2Ddiscrepancies,3Ddiscrepancies)amongthedifferentimplantsitesandthegenderofdentalstudentoperators.Thestudypopulationwas senior dental student volunteers. The outcomemeasureswerethepositionaldeviationsofimplantosteotomyandoperationtime.Itwasproposedthatimplantsitesmightinfluencethelearningofimplant

placementandmalesmightbebetterat learning implantnavigationthanfemales.Duetothenatureofthepilotstudy,anapriorisamplesizecalculationwasnotperformed.

3  | RESULTS

A total of 14 senior dental students were participated in thestudy.Thestudentparticipantswere recruited through their clini‐cal rotations intheoral&maxillofacialandperiodontologyclinics.DemographicdataarepresentedinTable1.Eachstudentplaced21implantswithinsixjaws(threesitesonthreereplicatemaxillae,foursitesonthreereplicatemandibles)usingdynamicguidednavigation.Atotalof294placedimplantswereassessedfortimeandaccuracy

F I G U R E 1  Schematicsdemonstratingthevideocameradetectionofthehandpiecepositionandjawpositionthroughmarkersortags

F I G U R E 2  Experimentalworkflow

4  |     GOLOB DEEB Et aL.

ofplacement.Statisticaldatawere summarised inTables2and3.Time to place a given implant depended on the attempt number(P <0.001), locationofthe implant (P =0.010),andmarginallyde‐pendedongender(P =0.061;Table4).

Timeimprovedsignificantlybetweenattempts1and2andthenplateaued(Table4,Figure6).Implantsplacedinareaofmaxillaryleftfirstmolartooksignificantlylongerthanothersites(Tukey'sadjustedP‐Value<0.05forall).Afteradjustingforattemptnumberandimplantsite,maleswere1.91minutesfasterthanfemales(P =0.061).Whenanalysingattemptsforeachjawindividually,timetoplaceanimplantimprovedsignificantlybytheattemptnumberformaxilla(P =0.003)andmandible(P <0.001).Maxillaimprovedacrossallthreeattempts;mandiblesecondandthirdattemptswerenotsignificantlydifferent(P =0.96;Figure7).Maxillaryleftfirstmolarimplantsitetookthelon‐gesttimeforimplantplacement(9.1minutes).Maxillaryrightcentralincisorsitehadtheshortesttimeof6.8minutes(Figure8).

Only implant locationwas significantly associatedwith differ‐ences inoverall 2D lateral deviation (P <0.001).Two‐dimensional

lateraldeviationdidnotimproveacrossattempts(P =0.513)ordifferbetweenmalesandfemales(P =0.345).Two‐dimensionallateralde‐viationwaslowestformaxillaryrightcentralincisorandhighestformandibularleftfirstmolar.(Table5andFigure9).

Theoverall3Dangulationoftheapexdependedonlyontheim‐plantsite(P <0.001),anddidnotimproveacrossattempt(P = 0.7840) orbetweenmales/females(P =0.372).Apex3Dangulationwasthelowestformaxillaryrightcentralincisorandthehighestformandib‐ularleftfirstmolar(1.37vs2.25,Table5).Theoverall3Dangulationof the implant depended on attempt (P = 0.0001) and marginally ongender (0.067),butdidnotdependon implantsite (P = 0.145). Overall3Dangulationimprovedsignificantlyfromattempt1‐2andthenplateaued(3.78vs2.87,Tukey'sadjustedP = 0.0264; Figure 10). Maleshadmarginallybetteraccuracy(2.41vs3.35,Tukey'sadjustedP = 0.067).

In addition to the objective outcomes, various subjective out‐comes were also measured through the survey completed aftereach attempt. The post‐experimental survey was done after all

F I G U R E 3   Automatic registration usingfiduciarymarkers/tags(arrows)

F I G U R E 4  Superimpositionofplannedandactuallyplacedimplantpositions

     |  5GOLOB DEEB Et aL.

experimentsweredoneforeachparticipant.Thesurveywasinten‐tionallyusedonlyaftertheexperimentstominimiseparticipant'sbiasby limited their knowledge about the system. Perceived bimanualstabilisationofthehandpieceimprovedsignificantlyacrossattempts(P =0.024;sevenaverageorbelowonfirstattemptvs0bysixthat‐tempt).Allstudents(whowereallright‐handed)perceivedleftsideas

morechallengingthanright,similartospeedandaccuracymeasures.Studentsreportedroughlyequalsplitbetweenwhichjawwaseasier:58%maxillavs42%mandible.Malesreportedmarginallybettervideogamingskillsthanfemales(P =0.10;43%stronglyagreevs0%),whichmayberelatedtothemarginaldifferencesinobjectiveoutcomes.

Since an a priori sample size calculation was not possible, aposthocpoweranalysiswasperformedtodeterminethedetect‐able difference for each outcome based on significance level of0.05,80%power,and14subjectswith21repeatedmeasures(21total implantsplaced).Using theobservedvariance and correla‐tionbetweenrepeatedmeasures(assumingcompoundsymmetricstructure),wehadpowertodetectadifferenceintimeof1.8min‐utes,2Dlateraldeviationof0.24mm,2Dverticalapexdeviationof0.21mm,overall3Dapexdeviationof0.21mmandoverall3Dangulation of 0.93°. These detectable differences are all largeenough to indicate clinical meaningfulness yet small enough tosuggest statistical validity.

4  | DISCUSSION

Itisinterestingtonotethatcomputer‐aided/assistedimplanttreat‐mentplanning,alsoknownasvirtualimplanttreatmentplanning,hasbeenusedwidely in thepredoctoraldental implanteducation.10,11 Dentaltreatmentmodelsimulationhasbeenwidelyusedindentaleducation.12However,itisimportanttopointoutthatthesesimula‐tionmodelsarehypotheticalandcannotbeusedinhumans.Dentalimplantnavigationsystemshoweverarebeingusedclinically.Onlyonestudydemonstratedthatdentalimplantnavigationsystemcanbeused toenhance implantplacement training in thepredoctoral

F I G U R E 5  Apexdeviation(3D)andentrylateraldeviation(2D)

TA B L E 1  Demographics

Mean SD

Age 26.1 1.77

n %

Sex

Male 7 50%

Female 7 50%

Dominanthand

Right 13 93%

Left 1 7%

Attempt (mean, SD)

1 2 3 4 5

Time(inmin) 10.5,3.93 7.5,3.58 7.9,4.05 7.4,3.52 6.4,2.88

Deviations

Lateral2D 1.2,0.64 1.2,0.54 1.2,0.72 1,0.6 1.1,0.62

Overall3Dapex 1.9,0.74 1.9,0.53 1.7,0.66 1.7,0.66 1.7,0.77

2DVerticalapex 1.1,0.51 1.3,0.58 1.2,0.54 1,0.63 1.0,0.62

Overall3Dangulation

3.7,2.33 2.9,1.52 2.8,1.80 2.6,1.63 2.5,1.46

TA B L E 2  Summarystatisticsbyattemptnumber

TA B L E 3   P‐ValuesforrepeatedmeasuresANOVAmodels

Attempt Location Gender

Time <0.001 0.010 0.061

Lateral2D 0.513 <0.001 0.345

Overall3Dapex 0.784 <0.001 0.372

2Dverticalapex 0.014 <0.001 0.767

Overall3Dangulation <0.001 0.1447 0.067

Numbersinboldrefertostatisticalsignificantvalues.

6  |     GOLOB DEEB Et aL.

dental students compared to conventional freehand osteotomypreparation.13While this particular study demonstrated that stu‐dentswhenusingnavigationsystemcanplaceimplantsinvitromoreaccuratelythanfreehandtechnique,theinvestigatorsdidnotlookat

thelearningcurveoriftherewasanydifferencebasedonimplantsitesorgenderofstudents.

This studywasperhapsoneof the first to demonstrate thatthere was a clear learning curve in training novice implant sur‐geonsusingcomputer‐guidedimplantnavigationsystem.There‐sults demonstrate three important findings. First, the hand skillwas improvedwithin the first three trial attempts and thenpla‐teaued out thereafter. This implies that using computer‐guidedtrainees with no implant placement experience would need atleast three trials before they would be comfortable in placingan implantwithoptimalpositioning in a timelymanner. Inothersurgical fields, such as laparoscopic surgery, endovascular sur‐geryor surgical endoscopy, computer‐guided simulationhasandcontinuestobeusedfortrainingandevaluationofprogressionofsurgicalaptitudeandcompetency.14‐17 Ithasbeendemonstratedthatintheseverydifferentsurgicalfields,surgicaladeptnessandlearning curve improved with training using computer‐guidedsimulation systems.14‐17Studentlearninginthisstudywasevalu‐atedthroughimprovedperformanceovertimeduringacquisitionofanewskill.Whileideallyacontrolgroupofotherinterventionshouldbeadded, the studyused the first implantplacement at‐temptasthebaselineorinternalcontrolgroup.Performancewasevaluated through improved time and accuracy of placement ofthe implant fixtures aided by guided navigation. In this study,the time required for osteotomy andplacement of implants sig‐nificantly improved between first and second attempt and con‐sequent attempts showed only slight continued improvementacrosstime.Thissupportsthelearningmodelthroughvirtualreal‐ityfeedbackandguidance.Sincevirtualfeedbackduringimplantplacementwithdynamicnavigationisenabledbycameratrackingwithtrackingtagsonmodelandhandpiece,itisimperativetokeepthoseunblockedinfullviewfortrackingarraysystem.

Second, the learning curves are different based on the im‐plant site. Implant site randomisation was used to control theexperimentalbias.Thedeviationsofthe implantswererecordedusingthesoftware inrealtimeandthereforetherewasnoneed

TA B L E 4  Factorsassociatedwithtimetoplaceimplant

Estimated* mean time (in min) SE P‐Value

Attempt <0.001

1 10.65 0.60 a

2 7.55 0.60 b

3 7.62 0.61 b

4 7.39 0.61 b

5 6.41 0.52 b

Location 0.010

3(Maxillaryrightfirstmolar)

8.24 0.63 a,b

8(Maxillarycen‐tral incisor)

7.19 0.63 a

14(Maxillaryleftfirstmolar)

9.51 0.63 b

19(Mandibularleftfirstmolar)

8.14 0.62 a

20(Mandibularleftsecondpremolar)

7.45 0.63 a

29(Mandibularrightsecondpremolar)

7.52 0.63 a

30(Mandibularrightfirstmolar)

7.41 0.63 a

Gender 0.061

Male 6.97 0.65 a

Female 8.88 0.65 b

*ResultsfromrepeatedmeasuresANOVAwithTukey'sadjustedposthoccomparisons;levelswiththesameletterwerenot.

F I G U R E 6  Averageimplantplacementtimebyattemptnumber

     |  7GOLOB DEEB Et aL.

F I G U R E 7  Averagetimebyattemptand jaw

F I G U R E 8  Averagetimebyimplantsite

TA B L E 5  Comparisonofdeviationbyimplantsite

Location

2D Lateral deviation (P<0.001) 2D vertical apex deviation (P<0.001) 3D apex deviation (P<0.001)

Mean* deviation SE Mean* deviation SE Mean* deviation SE

Maxillaryrightfirstmolar

1.01 0.10 a,b 1.08 0.09 a,b,c 1.67 0.10 b,c,d

Maxillaryrightcentral incisor

0.79 0.10 b 0.86 0.09 c 1.37 0.10 d

Maxillaryleftfirstpremolar

1.18 0.10 b,c 0.97 0.09 a,b,c 1.52 0.10 c,d

Mandibularleftfirstmolar

1.50 0.10 c 1.50 0.09 d 2.25 0.11 a

Mandibularleftsecondpremolar

1.41 0.10 c 1.29 0.09 a,d 2.08 0.11 a,b

Mandibularrightsecondpremolar

1.20 0.10 b,c 0.92 0.09 b,c 1.73 0.11 b,c,d

Mandibularrightfirstmolar

0.91 0.10 a,b 1.25 0.09 a,b,d 1.80 0.11 b,c

*Adjustedforattemptandgender;levelsjoinedbythesameletterwerenotstatisticallysignificantlydifferent;overall3Ddeviationdidnotdependonimplantsite.

8  |     GOLOB DEEB Et aL.

ofanexternalevaluator.Anteriorimplantsitesareeasiertolearnto place implants than the posterior ones. For this right‐handedgroup, the posterior left is the most difficult site to learn. Thislearningcurveappearstobesimilartostaticguidedsurgery.18 In thisstudysettings,thestentsusedonthemodelshavebeenpre‐fabricateduniformly,withtheJawTagplacedinonlyonepositioninrelationtothemodel,regardlessofthelocationofthesurgicalsite.Apossibleresultofthatwasthattherightsideandmandib‐ularplacementprovedtobemucheasier,possiblyexplainingthefindings fromactual time and accuracy aswell as perceiveddif‐ficulty related to certain implant sites. Depth of placement andadjacentteethdemandedtheuseofextendersinlowerleftquad‐rant.Extenderrendersthedrillmuchlongerandlessrigid,henceintroducing longerdistancefromthehandpiecetothetipof thedrillwhichmay affect calibration and accuracy. This could offera subjective explanation for self‐reported student perceptionthatimplantplacementinlowerleftquadrantwasmoredifficult.Objectively,thatperceptioncanbesupportedbylongertimesforplacementonlowerleftcomparedtolowerrightside.Despitetheimprovementsinoperativeorproceduraltime,therewasminimal

improvementindeviationsuggestingthatoncestudentsgetcom‐fortablewithproceduretheiraccuracydoesnotimproveanyfur‐thercorrespondingwithimprovedefficiencyandthattheguidednavigationsystemallowsforastableamountofaccuracy.

Third,therewereslightlearningadvantagesinthemaletraineespossiblyduetotheexperienceincomputergaming.Overall3Dan‐gulation,arguablythebestmeasureforoverallaccuracy, improvedovertimebutmostsignificantlybyattemptnumbertwoasstudentsgained more familiarity and implementations of skills respondingto navigation feedback information improved. In that perspective,videogamingappearstobebeneficialinadaptivelearningtointer‐activevirtualguidance.Itisnoteworthythatoverallmaleshadfasterimplant time placement and better overall 3D angulation but thismayberelatedtopriorvideogamingexperience.Asmalestudentsreportedgreaterexperiencewithvideogamesmayofferanexpla‐nationonwhymalesarebetter.19,20Theseriousgamingconcepthasbeenassociatedwithenhancementofskillprogressionandsimulatorvalidation in surgical trainees.19,20Somelimitationsofthestudyin‐cludethe limitationofpost‐experimentalsurvey,notraditional im‐plant trainingcontrol,andnoprior informationonpossible related

F I G U R E 9   Average deviation by attemptnumber

F I G U R E 1 0   Average deviation by implantsite

     |  9GOLOB DEEB Et aL.

skills such as computer gaming experience. It is also important toemphasisethatthestudyisapilotstudywithasmallsamplesizeinstudentswithlimitedknowledgeofimplantology.Futurelargerstudyshouldbedoneinvarietyofimplantsurgicalskillandknowledge.

Dynamicallyguidedsystemsmaypresentateachingtoolinearlydevelopmentofclinicalskillsinimplantplacementforthenoviceop‐erator.Thisinteractivemodelmayallowforthedevelopmentofneu‐ralpathwaysthroughbiofeedbackandmaybebeneficialinachievingmoreoptimal clinical results in theearlyphaseof surgical implanttraining.Noviceoperatorsoftenstrugglewithachievingcorrectdrillpositionincertainareasofthemouth,inparticularinposteriorsitesoppositetotheoperator'sdominanthand.Itshouldbekeptinmindthatclinicalscenarioscanbemorecomplexbeyondjusttheaccuracyofosteotomypreparationandsimulatedexperienceisnosubstituteforliveclinicalexperience.Futurestudiesshouldincludeexaminingtheroleofpreviouscomputergamingskillandtrainee'sgender,thelongerperiodoftrainingaswellasarandomisedcontrolledtrialtocomparethenavigationtechnologywithtraditionalimplantsurgerytraining.Itwillalsobeinterestingtoseewhetherthenavigationsys‐temcanbeusedtotrainstudentsforotherimplant‐relatedsurgicalproceduressuchridgeaugmentation,andsinusgrafting.Whilethisstudy showed a statistical significance in the improvement of im‐plantplacementusingnavigationsystem,itdidnotproveifthenav‐igationisbetterorworsethanconventionaltraining.Futurestudiesshouldincludethecomparisonbetweennavigationtrainingandcon‐ventionaltechnique.Theimplantplanningandissueofexperiencedvsinexperiencedsurgeonthatwereunfortunatelybeyondthescopethisstudyshouldalsobeaddressedinfuturestudies.

5  | CONCLUSION

Noviceoperatorsdemonstratesignificantimprovementofimplantplacement skills with dynamic navigation within three attempts.Thespeedandangulationdeviationimprovesignificantlywithinthefirstthreeattemptsandaresustained.Performanceformales,onaverage,wasmarginallybetterintimeandaccuracythanfemales.

ORCID

Sompop Bencharit https://orcid.org/0000‐0003‐1209‐9362

R E FE R E N C E S

1. BlockMS,EmeryRW,CullumDR,SheikhA.Implantplacementismore accurate using dynamic navigation. J Oral Maxillofacial Surg. 2017;75:1377‐1386.

2. TahmasebA,WismeijerD,CouckeW,DeksenW.Computertech‐nology application in surgical implant dentistry: a systematic re‐view. Int J Oral Maxillofacial Surg.2017;29:25‐42.

3. Scherer U, Stoetzer M, Ruecker M, Gellrich NC, von See C.Template‐guidedvsnon‐guideddrillinginsitepreparationofdentalimplants.Clin Oral Investig.2015;19:1339‐1346.

4. DeebGR,AllenRK,HallVP,WhitleyD3rd,LaskinDM,HowBS.Howaccurateare implant surgicalguidesproducedwithdesktopstereolithic3‐Dprinters?J Oral Maxillofacial Surg.2017;75:2559e1‐2559e8.

5. CassettaM,BellardiniM.Howmuchdoesexperienceinguidedim‐plantsurgeryplayaroleinaccuracy?Arandomizedcontrolledpilotstudy. Int J Oral Maxillofacial Surg.2017;46:922‐930.

6. BlockMS,EmeryR.StaticorDynamicnavigationforimplantplace‐ment‐choosing themethod of guidance. J Oral Maxillofacial Surg. 2016;74:269‐277.

7. Block MS, Emery RW, Lank K, Ryan J. Implant placement ac‐curacy using dynamic navigation. Int J Oral Maxillofac Implants. 2017;32:92‐99.

8. Somogyi‐GanssE,HolmesHI,JokstadA.Accuracyofanovelproto‐typedynamiccomputer‐assistedsurgerysystem.Clin Oral Implants Res.2015;26:882‐890.

9. VanAsscheN,VercruyssenM,CouckeW,TeughelsW,JacobsR,QuirynenM.Accuracyofcomputer‐aidedimplantplacement.Clin Oral Implants Res.2012;23:112‐123.

10. KiharaH,SunJ,SakaiM,NagaiS,DaSilvaJ.Asurveyofdentalim‐plantinstructioninpredoctoraldentalcurriculainNorthAmerica.J Dent Educ.2017;81:1085‐1090.

11. NkenkeE,VairaktarisE,BauersachsA,etal.Acceptanceofvirtualdentalimplantplanningsoftwareinanundergraduatecurriculum:apilotstudy.BMC Med Educ. 2012;12:90.

12. PerryS,BridgesSM,BurrowMF.Areviewoftheuseofsimulationin dental education. Simul Healthc.2015;10:31‐37.

13. CasapN,NadelS,TaraziE,WeissEI.Evaluationofanavigationsys‐temfordentalimplantationasatooltotrainnovicedentalpracti‐tioners. J Oral Maxillofac Surg.2011;69:2548‐2556.

14. Schiven MP, Jakimowicz J. The learning cure on the Xitact LS500 laparoscopy simulator: profiles of performance.Surg Endosc. 2004;18:121‐127.

15. AggarwalR,BlackSA,HanceJR,DarziA,CheshireNZ.Virtualreal‐itysimulationtrainingcanimproveinexperiencedsurgeons'endo‐vascularskills.Eur J Vasc Endovasc Surg.2006;31:588‐593.

16. AggarwalR,GrantcharovTP,EriksenJR,etal.Anevidence‐basedvirtual reality training program for novice laparoscopic surgeons.Ann Surg.2006;244:310‐314.

17. VerdaasdonkEG,StassenLP,SchijvenMP,DankelmanJ.ConstructvalidityandassessmentofthelearningcurvefortheSIMENDOen‐doscopicsimulator.Surg Endosc.2007;21:1406‐1412.

18. BencharitS,StaffenA,YeungM,WhitleyD3rd,LaskinDM,DeebGR. In vivo tooth‐supported implant surgical guides fabricatedwith desktop stereolithographic printers: fully guided surgery ismoreaccuratethanpartiallyguidedsurgery.J Oral Maxillofac Surg. 2018;76;1431‐1439.

19. Verdaasdonk EG, Dankelman J, Schijven MP, Lange JF, WentinkM, Strassen LP. Serious gaming and voluntary laparoscopic skillstraining: a multicenter study. Minim Invasive Ther Allied Technol. 2009;18:232‐238.

20. Winn J,HeeterC.Gaming,gender, and time:whomakes time toplay?Sex Roles.2009;61:1‐13.

How to cite this article:GolobDeebJ,BencharitS,CarricoCK,etal.Exploringtrainingdentalimplantplacementusingcomputer‐guidedimplantnavigationsystemforpredoctoralstudents:apilotstudy.Eur J Dent Educ. 2019;00:1–9. https://doi.org/10.1111/eje.12447