Managing the variability of biomechanical characteristics before the preliminary design stage of a medical device

Science Arts & Métiers (SAM)

is an open access repository that collects the work of Arts et Métiers Institute of

Technology researchers and makes it freely available over the web where possible.

This is an author-deposited version published in:

https://sam.ensam.eu

Handle ID: .

http://hdl.handle.net/10985/8508

To cite this version :

Michel MESNARD, Antonio RAMOS, Nicolas PERRY - Managing the variability of biomechanical

characteristics before the preliminary design stage of a medical device CIRP Annals

-Manufacturing Technology - Vol. 63, n°1, p.161-164 - 2014

Any correspondence concerning this service should be sent to the repository

Managing

the

variability

of

biomechanical

characteristics

before

the

preliminary

design

stage

of

a

medical

device

M.

Mesnard

a

,

A.

Ramos

b

,

N.

Perry

(2)

c,

*

a_{Universite´ deBordeaux,I2M,CNRSUMR5295,FR-33405Talence,France} b

UniversidadedeAveiro,DepartamentodeEngenhariaMecaˆnica,PT-3810-193Aveiro,Portugal

c

ArtsetMe´tiersParisTech,I2M,CNRSUMR5295,FR-33405Talence,France

1. Introduction

Whendevelopingmedicaldevices,theclinicalstudiesrequired andtheveryhighlevelofspecificationsforcertificationprocedures oftenrestrictresearchanddevelopmentdepartmentsto innova-tionsinvolvingalterationsanditerations.

Insteadofthisquasi-empiricalapproach,acompletestructured procedureshouldbeusedforaradicalbreakthroughinnovation whendesigninganarticularMD(prosthesis,implant).Functional analysiscanthenbeappliedtosearchforinnovativeconceptsand producebehaviouralmodelsforthejointinitsnaturalstateand/or when equipped with a prosthesis. Moreau–Gaudry and Pazart proposedaframeworkforatechnologicalinnovationdevelopment forhealthapplication,basedon: concept-research-tests-product-treatment,cycle[1].

Wepresentstudiescarriedoutbeforethepreliminarydesign stageofaninnovativetemporomandibularjoint(TMJ)prosthesis. Thispaperdescribesthemanagementandintegrationof anatomi-calandfunctionalvariabilitiesusingamodulardesignapproach. TocharacterizeahealthyTMJandestablishdesigncriteria,joint displacements were defined and quantified experimentally; results produced some very wide intra- and inter-individual variations.

Ouraimisnottocontrolthesevariationsbutrathertodesigna prosthesisthatcouldcarryoutitsfunctionswhileincorporating natural or acquired articular fluctuations, displacements and geometry.

Theexampledescribedherepresentsanexperimental statisti-calanalysiswhichrecordsthedistributionofthetemporalslope

angle.Using thisdistributionwewereabletocreateamodular component, selecting three to five values for the angle being defined.

The difficulties involved in the production of a made-to-measureprosthesisandthelimiteddistributionoftheTMJguide thedesignertowardsamodularsolution.

During the design phase of an innovative TMJ prosthesis, functionalanalysistechniqueshighlightedtheneedfora prelimi-nary study to characterize the morphological and functional anatomy ofthejointin ordertodefineassessmentcriteria and quantifyacceptancerequirements[2].Thisstudyfocusesonthe geometricaldefinitionforthedesignoftheTMJkeycharacteristics. Thereislittleconsiderationtobio-materialdesignand optimiza-tion(surfacecoatingortextures)asreferbyRamsdenet al.[3]. Moreover,theadditivemanufacturingopportunitiesonshapeand materialpossibilities,ashighlightedbyBartoloet al.[4]aretoday outofthescope,butmaybecomesolutionstofacetheproblemof variabilityillustratedinthispaper.

Different techniques exist to identify and re-create bones morphologyandshapesbasedonX-raysormagneticresonance imaging(MRI)[5].

Inthisstudywedevelopedareverseidentificationbasedonthe patientmandibularmobilitymeasurementsinordertocalculate theTMJgeometricalparametersneededfor thedesign.Specific metrologytechniquesweredevelopedtoquantifybiomechanical characteristicsexperimentallysuchasdisplacements,actions,etc. TMJdisplacementsfromthepointofarticularcontact(Fig.1)were quantified by stereophotogrammetry and muscle efforts were assessedbyelectromyographyandMRI[6].

Resultsshowedconsiderableintra-andinter-individual varia-tions.Thisstudydescribeshowthesevariationsweremanagedand integratedusingamodulardesignapproach.

Keywords: Designmethod Biomedical Modulardesign

ABSTRACT

Theveryhighlevelofrequirementsforcertificationproceduresoftenlimitresearchanddevelopment departmentstoinnovateusingincrementsanditerationsduringthedesignprocessformedicaldevices (MD).Insteadofthissemi-empiricalapproach,astructuredprocedure,abreakthroughinnovationshould beusedwhendesigninganarticularMD(prosthesis,implant).Thesearchforconceptscanbebasedon functionalanalysisandproducingbehaviouralmodelsofthejointinitsnaturalstateand/orequipped withtheprosthesis.Thispapershowshowanatomicalvariablescanbemanagedandintegratedusinga modulardesignapproach.

* Correspondingauthor.

2. TMJprostheses

Withintheworldpopulationasawhole,WolfordandMehra pointoutthattheimplantationofaTMJprosthesisconcernsfewer than five thousand patients every year suffering from acute arthritisandfibrosis[7].

Occlusioncorrectionsclearlygivethepatientbothphysicaland psychological comfort and comparison of actual prostheses underlinesthattheinter-incisoropeningcanreach20ml. Specu-landalso pointsout differentcasesof failuresthat can require additionalsurgicalprocedures:

-Inflammation occurs with persistence of pain revealing an imperfect osseointegration of the screws accompanied by implantmicro-displacementswithrespecttothebone, -Prosthesisfracture(rare)happensnearascreworinavariationof

thesectionwherestressestendtoconcentrate,

-Bonefractureseemstobetheresultofaninsufficientnumberof screwsoratoohighintensityofloadsonthecondyle[8].

Thetwoprostheticmodelscurrentlyavailable(Fig. 2)requirea veryinvasiveapproach,andmajorboneresection.Theygeneratea riskofanteriororposteriordislocationofthejointandreducethe amplitude of articular displacement in translation

D

. Yet this displacementisnecessarytoestablishmouthopeningandgivean inter-incisordistancewhichiscomfortableforthepatient.

Ramoset al.havemeasuredthemicro-mobilitybetweenthe implantsurfaceandboneandshowedthatitdecreaseswhenusing ananatomicalplate.Largemicro-motionscaninducethe forma-tion of fibrous tissue between bone and implant and produce instability.Inthesametime,thelowerstiffnessoftheanatomical implant can reduce stress shielding effects and produce a favourablebiomechanicalenvironmentforosseo-integration[9]. The dual goal of an innovation will now be to reduce the cumbersome nature of the temporal element and modify the

mandibularelementfixationsoastoresemblenaturalopeningand occlusion.

In ordertocarryout a conceptsearchand thenprovidethe elementdimensions,thetemporalgeometrymust be character-izedtodeterminethevalueofthetemporalslopeangle

a

(Fig. 2) andprovideinformationonindividualvariations.

3. Geometryofarticularsurfaces

A repeatability study was carried out on the experimental protocol to characterize TMJ articular displacements. This was thenimplementedandvalidatedusingtwovolunteers.

3.1. Protocolandstereophotogrammetry

Tostudythedisplacementsofthemandibleinrelationtothe skullinourvolunteer,referencepointswerelocatedonthelower jawandonthemaxilla.Becauseprecisionwasessential,reference points were not located on the skin, as these can move over adjacentareasofbone[10].

Twogutters, interdependent ofthe two dental arches,were moulded from impressions taken from the volunteer using orthodonticmaterials.Arigidmetalsnapringwasinsertedinto eachgutterandaplatewasattachedtothering(Fig.3).Thecranial referencepointontheupperplatewasdefinedusingthreetarget points, with point H representing the barycentre, the lower referencepointwasassociatedtopointBonthemandibularplate inthesameway.

Inthesagittalsymmetricalplanetheprojectedcentreofthe condylesisinpositionC0initiallyandiswrittenCduringopening

movement.VectorsC0HandBC,whichareknownmorphological

characteristics, remain constant at the cranial and mandibular referencepointsrespectively.

During the open–close movementand recording, vectorC0C

whichrepresentsthedisplacementoftheprojectedcentreofthe condylescanthenbecalculatedbythevectorsumC0C,

C0C¼C0HþHBþBC where;

vectorHBisobtainedbystereophotogrammetry[11].

The open–close movement (Fig. 3), facilitated by the TMJ meniscus,istheresultofslidingoncontactandthesimultaneous rotationoftheangle

u

ofthemandibleinrelationtotheskull. 3.2. Repeatability,uncertainty

To analyze the repeatability of the protocol, theupper and lowerjawplateswerefixedtoabase.Therelativepositionsofthe targets,andthedummyreferencepointsC0andC(Fig. 3)were

thusconstant.

Fig.1.TMJandvisualizationofdisplacements.

Fig.2.Prosthesisanddefinitionofslopeangle.

Theequipmentwasdisplacedmanuallywithinthefieldofthe cameras;theamplitudeoftheimposeddisplacementswasvery muchgreaterthantheamplitudeoftheheadmovementsoverall. Datawererecordedandprocessedasdescribedabove.

Thus displacements calculated from point C, the imagined centreoftheTMJ,inrelationtopointC0,quantifiedtheuncertainty

ofthemeasurementassociatedwiththeoveralldisplacementof theplatesinthefilmedarea.

Theabsolutevalueoferrorremainedlessthan0.7mmwhile displacementfrompointCcouldreach20mmatmaximalopening. Thevalueoftheratio20.7/20(7%)confirmedthatthemethod wasvalid.

3.3. Characterizationanddesignvariables

MajorobjectivesinalloplasticTMJreconstructionfor improve-ment of life quality are pain relief and a better control of the mandibularkinematics[12].Thiscontrolofmandibularkinematics can be gauged in particular by a significant increase in inter-incisivedistanceduringmaximumopening.

Designvariablesinherentinthegeometriccharacteristicsofthe TMJdeterminemaximuminter-incisivedistance.Thisdistanceis influenced simultaneously by length of the slope

D

, angle of inclination

a

andamplitudeofrotation

u

.

Astherotation

u

andthez-translationofpointCrepresentthe two main displacements during an open–close movement, the relationbetweenrotatorydisplacementandtranslatory displace-ment was defined by a coefficient named the ‘‘preponderance coefficient’’asfollows:

Cp¼ Rotation

u

z-translation Tz where;

u

wasexpressedin(8)andTzin(mm).Then,this coefficientCp quantifiesa ratio which is characteristic of thesubject studied [13,14].

4. Variationsincharacteristics

Duringtheconceptsearch,theaimisnottocontrolindividual variationsinbiomechanicalcharacteristics,butrather,basedon quantification,todesignandoptimizeaprosthesisthatcancarry outitsfunctionswhileatthesametimeincorporatingintra-and inter-individualfluctuations.

4.1. Intra-individualvariations

Intra-individualvariabilitymayresultfromarangeofcauses: temporal (structural modification of tissue due to ageing), pathological, behavioural (environment, fitness, experimental conditions,etc.).

In a healthy volunteer, repeated recordings (three to five) enable us to assess the influence of changes in behaviour, to establish a significant value for the inter-incisive maximum openingdistance,displacementsofpoint Cinthesagittalplane andhencethenaturaltemporalslopeangle

a

(Fig. 4).

4.2. Inter-individualvariations

Inter-individualvariability,ontheotherhand,islinkedonlyto anatomical-physiologicalfactorsspecifictotheindividual volun-teers (overall geometry and local distribution of biological tissue...).

Arepresentativesampleofthirty-twosubjectswasstudiedwho couldpotentiallybeconcernedlaterbyaTMJprosthesisimplant. Pathologicalcaseswereexcludedfromthis study;foursubjects withmajorasymmetriesorpathologieswerenotselected.

TheCpcoefficientwasdeterminedforeachstandardvolunteer. Itwasacontinuousrandomvariablewhosemeanvalue(2.07)and squaredeviation(0.77)weredeterminedin thesample [13,14].

This coefficient distribution could be studied by applying a Laplace–Gaussfunctionforeachxisubjectvalue,bytherelation:

fðxiÞ¼ 1

s

p effiffiffiffi

p

1=2ðxim=sÞ2

Thevariablewasdeterminedrelativetothemaximalopenjaw amplitudeandfollowedanormaldistribution(Fig. 5).Theresults discriminategroupspresentingweakinter-individualvariationsin relationtothecollectivevalues[13].

5. Designcriteria

Table1showsthevaluescalculatedfortheCpcoefficientand dimensionsoftemporalslope,length

D

,slopeangle

a

andtheir square deviations. These two characteristic slope elements determinethechoiceofprosthesisormodulesduringsurgery.

5.1. Temporalslope

Three groups were formed from the mean value of Cp by standard deviation. Three well-individualized values of the temporal slope angle can then bedistinguished, corresponding to each of the three groups as well as weak inter-individual variations.

Fig.4.PointCdisplacementsinthesagittalplaneforonevolunteer.

Fig.5.DistributionofCpcoefficientamong32volunteers.

Table1

DisplacementandslopeangleaversuscoefficientCp.

Group 1-Tr. 2-Mix. 3-Rot. Cp [1.02,ms] [ms,m+s] [m+s,3.76]

Cpmean 1.0 2.1 3.6

nb.subjects 7 17 4

D(mm) 11.4(1.2) 10.5(1.0) 6.1(1.1)

Thus a temporal slope can be correlated with a kinematic characteristic.

For group3 (4volunteers), mouth openingis obtained by a quasi-pure rotation around the condylar axis and a short z-translation.Forgroups1(7volunteers)and2(17volunteers),the openingmovementisintwo phases:first,concomitantrotation andtranslationofpointConthetemporalslopeandthen quasi-purerotation.Duringthissecondphase,majordemandsaremade onthedisc-condylesystem,whichisvirtuallystationaryandshort forgroup1.

5.2. Partialdefinitionoftheimplant

In ordertoget closeto thepatient’s specific geometricand kinematiccharacteristics,fivemodelsofthetemporalpartofthe prosthesis are defined.Three correspond tothe groups formed previously; two complementary models correspond to cases locatedattheintervalboundaries(seeFig. 6).

6. Conclusions

Thereisnevertranslatorydisplacementwhichisnotcombined with rotatory displacement. This finding makes it possible to definea characterizationcoefficient,toanalyze inter-individual variationsandthustodiscriminatebetweenindividualsinorderto determinegroupspresentingcommonkinematicaland geometri-calcharacteristics.

The disc-condyle displacements associated with open jaw amplitudecanbeanalyzedineachanatomicalzoneandforeach group. Thus, the correlation between the kinematic character definedbyCpandthetrajectoryrelatedtothegeometriesofthe temporal-bonefacetscanbeassessed.Eachgroupiscorrelatedtoa geometricalmodelthatgeneratesaspecifictrajectory.Theresults of the experimental study determine theangle

a

defining the trajectoryof thecondyle centre along thetemporal slope with respecttotheCamper’splane.

ThelimiteduseofTMJprosthesescombinedwithdifficultiesin producing a made-to-measure component tend to guide the

designertowardsamodularsolution.Foranoperation,aninitial selection can be made by preoperative scan which is then confirmedwhen theoperation is in progress.By providing the surgeonwitharangeoffivetemporalmodules,theslopeanglethat ismostsimilartothatofthepatientbeingtreatedcanthenbe selected.

Theadditivemanufacturingand3Dprintingtechnique devel-opedforrapidprototypinghasalreadybeenusedsuccessfullyto producefinishedpartsinindustriessuchasaerospace,dentistry, etc. It will eventually be possible to produce a prosthesis componentthatisimplantable,biocompatibleandtailor-made. Acknowledgments

Thisstudyhasbeenrealizedunderthetwojointactionprojects PESSOA14630YAandPTDC/EME-PME/112977.

References

[1]Moreau-GaudryA,PazartL(2010)De´veloppementd’uneinnovation techno-logique en sante´: Le cycle Concept-Recherche-Essais-Produit-Soins. IRBM 31(1):12–21.

[2]MesnardM,RamosA(2013)TowardsaRigorousApproachtoDesigninga TemporomandibularJointProsthesis.FromClinicalChallengetoNumerical Prototype.ProcediaCIRP5:141–146.

[3]RamsdenJ,AllenDM,StephensonDJ,AlcockJR,PeggsGN,FullerG,GochG (2007)TheDesignandManufactureofBiomedicalSurfaces.CIRPAnnals– ManufacturingTechnology56(2):687–711.

[4]BartoloP,KruthJP,SilvacJ,LevyG,RajurkarK,MitsuishiM,CiuranaJ,LeuM (2012)BiomedicalProductionofImplantsbyAdditiveElectro-chemicaland PhysicalProcesses.CIRPAnnals–ManufacturingTechnology61(2):635–655.

[5]MajstorovicaV,TrajanovicbM,VitkovicbN,StojkovicbM (2013)Reverse EngineeringofHumanBonesbyUsingMethodofAnatomicalFeatures.CIRP Annals–ManufacturingTechnology62(1):167–170.

[6]MesnardM,RamosA(2012)BiomechanicalCharacterizationandModellingof NaturalandAlloplasticHumanTemporomandibularJoint.Technologiesfor Medical Sciences, Lecture Notesin Computational Vision and Biomechanics, 39–66.

[7]WolfordLM,MehraP(2000)Custom-madeTotalJointProsthesesfor Tempo-romandibularJointReconstruction.ProceedingsBaylorUniversityMedical Cen-ter12(2):135–138.

[8]SpeculandB(2009)CurrentStatusofReplacementoftheTemporomandibular JointintheUnitedKingdom.BritishJournalofOral&MaxillofacialSurgery 47(1):37–41.

[9]RamosA,CompletoA,RelvasC,MesnardM,SimoesJA(2011)Straight, Semi-anatomicandAnatomicTMJImplants:TheInfluenceofCondylarGeometry andBoneFixationScrews.JournalofCranio-MaxillofacialSurgery39(5):343– 350.

[10]ChiariL,etal(2005)HumanMovementAnalysisusingStereo Photogramme-try–InstrumentalErrors.GaitandPosture21:197–211.

[11]Abdel-AzizYI,KararaHM(1971)DirectLinearTransformationfrom Compar-atorCo-ordinatesintoObjectSpaceCo-ordinatesinCloseRange Photogram-metry.ProceedingsoftheSymposiumonCloseRangePhotogrammetry,1–18.

[12]LinsenS,ReichR,TeschkeM(2012)PressurePainThresholdandOral Health-relatedQualityofLife ImplicationsofPatientswithAlloplastic Temporo-mandibular Joint Replacement. Journal of Oral and Maxillofacial Surgery 70:2531–2542.

[13]CoutantJC,MesnardM,MorlierJ,BalluA,CidM(2008)Discriminationof ObjectiveKinematicCharactersinTemporomandibularJointDisplacements. ArchivesofOralBiology53(5):453–461.

[14]MesnardM,CoutantJC,AounM,MorlierJ,CidM,CaixP(2012)Relationships BetweenGeometryandKinematicCharacteristicsintheTemporomandibular Joint. Computer Methods in Biomechanics and Biomedical Engineering 15(4):393–400.