CLlN'iCAL ORTHOPAEDlCS AND RELATED RESEARCH
`Number 354, PP 28w88
`© 1993 Lipp‘mcotl Williams & Wilkins
`
`Computer Assisted Orthopaedic
`Surgery With Image Based
`Individual Templates
`
`Klaus Rade‘rmacher, Dipl-Ing*; Frank Partheine, Dipl-Ing*;
`Marc Anton, Dipl-Ing**; Andreas Zimolong, Dipl-Ingl;
`,Gii'nther Kaspers, MD“; Giinter Ran, PI;D*;
`V
`and Hans~Walter Stazzdte, MD**
`
`in computer assisted
`Recent dovelopmems
`surgery offer promising solutions for the trans,-
`lation :of the high acwracy of the preoperative
`imagingand planning into precise intraopérw
`ti‘ve surgery. Broad clinical appliéation is him
`dared by high costs, additional time during in-
`tervention, problems of intraoperaiive man and
`machine into:action, and the spatially con»
`strained arrangement of additional Equipment
`within the operating theater. An alternative
`'tochniquo'for computerizod tomographic image
`basal preoperative three-dimensional gunning
`and precise surgery on bone structures using in»
`divifluai templates has lie-on tievel‘opofit For the
`preoperative customization of these mechanical
`tool guides, a desktop computer controlled
`milling device is. used 'as a three-dimensional
`printer to mold the shape. of small reference an
`cos of the bone 'suufaw automatioally into the
`boxiy of the template. Thus, theplannod position
`and orientation of the tool' guide in spatial rela-
`tion to bone is stated in a struCtural way and can
`be reproduced 'intraoporatively by adjusting the
`
`
`From the *Heimholt: institute for Biomedical Engl~
`nearing; Aachen University of Technology, Aachen,
`Germany; and ”Department for Orthopedic Surgery,
`District Hospital Marienhohe Wflrselcn Germany
`Suppottéxl in part by the Earopoan Commissionto the
`framework of the Tciemmics Applications Program:
`(TAP) project
`IGOS (imago Guided Gribopcdic
`Surgery) project number HClOZfiHC
`Reprints requests to Klaus Radetmacher, Dipl-Ing-,
`Helmholtz lustiiut for Biomedical Engineering, Aachen
`University of Technology, Pauwelsstrasse 2'0, D620”:
`Aachen, Germany
`'
`
`position of the customized contact faces of the
`template until the location of exact fit to the
`bone is fount’i. No additional computerized
`équlpment or time is needed during surgery.
`The feasibility of this approach has been shown
`in spine, hip, and knee surgery, and it has been
`applied clinically for pelvic repositioning os~
`mommies inzacetabular dyspiasia therapy.
`
`Recent research activities in the area of com
`purer assistod surgery have concentrated on
`tho introductidn of additional sensor and robot
`based guiding systems into the operating the»
`ater to ooable admoate computer assisted
`transform] of the high accuraCy of preopera—
`tivo imaging and planning to prociso surgery
`Different promising solutions have boon do-
`volopod concoming the related problems of
`multimodal information processing and regis~
`nation. safety and sensor concepts, and ado
`quato control
`stwttegi$5.539,22 Nevertheless,
`problems still remain such as the time needed
`,imraoperafivoly for the interaction with addi~
`tionai technical system components, for the
`additional intraoperative registration of bone
`structures,
`the spatial arrangement oi" dis—
`plays, sensors and robot systems within, the
`operating room, the overall costs of the sensor
`or robot based systems, and mismatches rev
`gaming ergonomic design aspects of cogni~
`lion and manual control." An exact medical,
`technical, and ergonomic analysis is neces~
`
`WMT 1006-1
`
`WMT 1006-1
`
`

`

`Number 354
`September-.1998
`sary to identify during which phases and se-
`quences of intraoperative work additional
`guiding systems or even robotic assistance
`contributing a specific complementaiy .func»
`tion would be required.
`In this context, the task sequences of dif—
`ferent conventional orthopaedic. interven»
`ti‘ons Were investigatedfl16 such as total hip
`and total knee replacement surgery being
`typical applications for computer assisted
`Surgery systems9.19 One result of these in»
`vestigatinns was that observable activities of
`direct orientation6 only took. approximately
`2% to 4% of the overall operating tinielf all
`activities are taken into account that have
`some connection with exact
`threeédimem
`
`Sional positioning, their share of total inter~
`ventinn time is
`approximately 10% to
`15%)”6 Hence, the potential tn shorten op“
`erating. time and the related time. slot. for the
`introduction of computer assisted guiding
`systems, seems lobe limited; Prolongatien of
`the Operating time by additional intraopera~
`tive tasks and interactions with additional in-
`
`traoperative equipment can be tolerated only
`to a limited degree because of medical, orga—
`nizational, and economic reasons. In every
`case, it would have to be justified bya‘signifis
`cant. enhancement of the therapeutic. outw
`come. Moreover,
`the costs of additional
`equipment must be balanced against the rew
`quirement to deliver higher quality treatment
`to a'iarger number of‘patients.
`The goal of the work described here was
`todevelop a relatively simple, low cost 5.0111»
`tion that facilitates exact, safe, anti fast im-
`plementation of planned surgery on bone
`situctures,. eliminates the need for continua-l
`rad‘i’egraphic monitoring; and, avoids over-
`burdening surgery with complex equipment
`and time consuming procedures
`I
`
`PRINCIPLE 0F INDIVIDUAL
`TEMPLATES
`
`In orthopaedic surgery standard, template sys»
`, terns are familiar technical means-to guide
`drills and saws in total knee replacement.
`
`Surgery With l'mageyBased. Individual Templates
`
`29.
`
`However, the design of these tool guides is
`based on averaged anatomic geometries. The
`positioning of the template onthe bone bears
`no precise relationship to the position defined
`by individual preoperative planning.
`Essentially, the missing information is the
`precise spatial correspondence between the
`individual bone structure in situ and. the in—
`
`tended position of the toolgu‘ides.
`The author‘s
`investigated a means of
`adding this missing information to the class
`sic templates by- providing. shape based phys—
`ical matching between the reference surface
`of the individual bone and the reference sup
`.face of the computer based model. This in—
`formation is incorporated into an individual
`'tetnplate.1'2“1.6
`Individual templates are customized on the
`basis of three»d3irnensi0nal reconstructions of
`the bone structures extracted from computer-
`ized tomographic (CT) image data in accor—
`dance with individual preoperative surgical
`planning. For preoperative customization, a
`low cost desktop milling machine is used, as a
`threenvdimens‘ienal printer to mold the shape of
`a small reference area on the individual, bone
`automatically into the template. By thistneans
`the,planned position and orientation ofthe tool
`guide in spatial relation to the bone is stored in
`a; stmctnral way andcan be teproduced in site
`adjusting the position of the contact faces of
`the template until they fit exactly on the bone
`(Fig 2). Neither iterative time consuming work
`under tadiographio control or registration pm»
`cedures, nor any additional computerized
`equipment is needn‘t intraoperatively. Mecham
`ice! guides for (hills, saws, Chisels, or milling
`tools are adaptable or integratedinto these. indi~
`vidual templates in predefmed positions for
`different types of interventions. Moreover,in»
`dividual templates also canbe used forfixat-ion
`of a referenee base for stamiaxd tool guides or
`other devices in a defined position on bone;
`The individual, templates can be autoclaved
`and delivered to the operating room with the
`conventional instrument sets. The feasibility
`and adaptability of this, approach has been
`shown on anatomic models andin cadaver tests
`
`WMT 1006-2
`
`WMT 1006-2
`
`

`

`Clinical Oithopaedics
`and Related Research
`Radermacher at al
`.30
`
`
`3i HPRNClPLFOFlNDlViDUAL TEMPLATES ‘ Z.
`
`ix; {drill guide
`
`.
`
`
`
`Fig 1 Padid'e screw placement: (A) the principle
`of individual templates baseo on preoperative C?
`imaging and computer based planning; (3) com;
`outer assisted planning of a podicis screw place‘
`ment with the DISCS planning aysiemf- (C)
`interactive specification of the refemnce contam-
`taco; (D) iormclosod lniraoperative fitting of the
`iem'piate; and {E} radiographio control of
`the
`placement oi 5mm rods in {he pedicles without
`any perioration‘in situ
`
`for various applications-Almflé Results of in
`vim; studies on accuraoy- have been reported in
`previous'pilblioafionsfimfi
`Among the applications of this technique-
`are pefliclo screw placement (especially in sow
`liosis
`therapy; Fig 3};
`reposgiziomng
`0&-
`ieotomies in Spine surgery; puncture of a cystic
`cavity in the; femoral head; intemochanteric
`repositioning osicotomy; initial reference QS~
`teo‘tomies. for total liner: repiacemeni
`(espe—
`cially in. the case of pathologic defonnations};
`pedanetabular repositioning ostootomies; open
`door decompression in.
`the Cervical spine;
`transcorpoml decompression in the cervical
`Spine; and decomprossion in the lumbal spine,
`
`To depict ih’e spootrum and. potential of_
`various implementations of {he principle. of
`individual templates, four typical exampies
`will be, described in more detail, with ompha~
`sis on the clinical application of triple repo—
`sitioning osteotomies for the treatment of ac—
`etabular djysplasiafi‘ziMAS-m22
`
`.PEDICLE SCREW PLACEMENT
`
`Tho selection of a podicle screw of the. appro-
`priate lengthrand caliber and its accurate fixa-
`tion in the cortical bone of the. pedicles and the-
`vemab‘ra‘l body is essential for good anchoring.-
`Perforations are me major specific complim-
`tion of pediclescrew placements; implying a
`high risk of bone weakening or lesions of the
`spinal cord nerve roots, or blood vesscis37 To
`monitor
`the placement of
`the
`piloting
`Kirsohner wire and finally the screw as many
`as four to five radiographs in diffeient planes
`are recommended per screw placement} ”She.
`auihois selected this application for their initial
`investigations of {he principle of inciividual
`templates”'3 (Fig .1) Human anatomic speci-
`mens of lumbar Spinos were scanned with CT
`(slices. 2~mm thick and 2mm again). The im
`age data were. transferred fio the, personal oom—~
`puter based. DISCS planning workstation
`(Gamma mbH, Aachen, Gonnany}. Based on
`ihreodimensional accumulations auiomalj~
`cally provided by the system, the surgcon se~
`lects «an approgriato screw and defines itsiopti
`mal glacemant (Fig {B}. The position and
`orientation of the related drill, goiéo .lhen is
`specified: and can he incomrated into the indi«
`vidual temolate. To provide Shape based intra—
`oparative matching, small referoncc contact
`faces to the Vertabral bone have to be specified
`on the display in thervioiniiy of the; transverse
`process, the arch, or the spinous procoss. To
`this end, the surgoon interactively selects and
`positions an appropriate templatc, within the
`three~dimensional View of the vertebral bone
`structure (Fig 1C). The surgeon is supported by
`the systom, which automatioally constrains the
`positioning to the {inaction of {he definedbore
`axis and ovalllates this Quality of the contact,
`
`WMT 1006-3
`
`

`

`Number 854.
`
`SurgeryWilh Image Based Individual Templates
`September, 1998
`
`31
`
`
`
`Fig 2A,-B. Total knee arthroplasty: (A) laboratory- investigation on a plastic hone model (1'): individ—
`ual template guiding the reference osteotomy- (a) in tibial. bone, optional fixation with a bone pin (4);
`(,8) Customized reference contact taco. (5)1and copying profile (6) limiting cutting depth (7) to the dor~
`sat contour (6), of tibial bone.
`
`faco..Aftorward,rtho system automatically gem
`exams the. manufacturing program and Gus-
`tomizes the semifinishod. template with its into~
`grated drilling guides.
`Intraoperatt'Vely,
`the
`defined gosition of the: bore is reproduced by
`placing the self locating template when: it fits
`exactly on the bone. Additionally, an optional
`radiographie manual of the. position of the:
`stainless- steel {killing guide is possible, 136*
`cause the plastic maiotial of the template body
`is radiolucont. In two cadaver studios, in which
`
`one maceraled sgecimon and. one specimen
`with soft tissue were used, 5~mm horas, have
`been reproduced exactly according to the pre-
`operative planning without any perforation it;
`sttu.
`
`TOTAL KNEE ARTHRQPLASTY
`
`In total knee arthroglasty accurato placoment
`of implant components. with respect to the
`individual mechanical axis of the leg is es,~
`sential. Conventionally,modular mechanical
`(lettings. corrosponding to the‘intrin'sle shape
`of the, implant componoms am used to guide
`the os‘teotomies- and bows for thoprepataiion'
`of the, implant’s seat.- By- motlnting these
`conventional tool guide systems on an indi-
`vidual 'tomplato as a basic customized. refer-
`ence, it is possible to reproduce the p‘tt’ooper~
`atlvely ‘plannotl posltion exactly even in the
`
`case of sevotely deformed bone. Mommas,
`for proser‘vationof the posterior cruciate lig—
`aments and the nerves and vessels inthe holv
`
`low of the knee, not'onl’y' the reference sup
`face of the: bone but also .3 copying surface
`limiting the cutting depth to tho dorsal com
`tour of tho tibia can be molded into the tem-
`
`plate (Fig ;23). The geometry of the saw has
`to Abok-noWn and a calibrated copying cam
`has to be mounted on the conventional Saw.
`Figure 2 shows a feasibility study with a
`CT image based indiyidual template for the
`.reference_ tibial out for total knee replacement
`on a plastic bone model.15 The geometry of
`the cut with its position, orientation, andlimi~
`nations was; planned on the basis of CT images
`(slices 24mm thick and 2mm apart). In addi»
`tion, topogtams could be» used. to identify the
`bone; axis. A conventional saw guide: can be
`mounted on the individual. template, :Whléh
`serves as a reference base :er subsequent
`work on the‘bone. The template hasfbeen ens;
`tomized in the areas :of‘ the reference surface
`
`the individual copying yrofile. corre~
`and,
`sponding to the dorsal contour of the tibial
`bone within the out plane. The accuracy of- the.
`reproduction was measured: directly (on the
`bone model using a conventional precision
`goniomotor and 3 cannot gauge. The prede~
`fmod-cut plane and the position of'the copying
`profile limiting the cutting depth were repro-
`
`WMT 1006-4
`
`WMT 1006-4
`
`

`

`
`32 , Radermaeher et‘ at
`
`Clinical Orthopaedics.
`and Related Research
`
`duced with an accuracy better than 1 mm in
`all directions and 1° inclination in the sagittal
`and transverso planes.
`
`OPEN DOOR DECOMPRESSION IN
`CERVICAL SPINE
`
`The authors investigated the possibility of exp
`plying this principle to open door decompres-
`sion in» the cervical spine, which involves
`even more delicate constraints on the required
`accuracy of imaging, planning, and imple~
`meniscionJ‘tel5 The task is to mill away one
`side of the lamina completely and to preserve
`the anterior cortical layer on the other side,
`which acts as a hinge for. the dorsal open door
`laminoplasty.‘ Figure 3 shows a corresponding
`design-of an individual template and the le~
`ning and intraoperatiye implementation in the,
`framework of la cadaver studyflii"! Applying
`the principle of copying profiles, a miniatun
`sized copying: cam corresponding to. the geom~
`.etry of the. milling tool was mounted on a
`standard micromilling machine, with a sliding
`jig piste to guide flie‘micromill perpendicular
`to the template. A boats with a defined depth
`was provided in the template» for a final on site
`calibration of ' the length of the milling tool.
`with respect to the copying cam Then, both
`laminetomies were performed within less
`than 19 minutes, subsequently removing sev»
`era} layers of bone to less then 1 mm on the
`left side and approximately 1 mm on the right
`side The mieromillin'g tool was guided by the
`copying profile, leaving the anterior cortical
`hinge of the right lamina intact. It perforated.
`only slightly at two. points on the left; side
`without damaging the dura'(Fig BEE-F).
`
`TRELE OSTEOTQMY on PELVIC
`some non THE TREATMENT on A
`ovsseasrrc HIP JOINT
`
`Periacetabular triple osteotomy for the. treat,~
`ment‘ of hip dysplasia was chosen asthe first
`exemplary clinical application of individual
`templates.Wt“8 in acetabular dysplasia the-
`taslc is to enlarge the weightbearing part of the.
`
`acetabulum covering the femoral head to re~
`duce- the pressure on this» area to physiologic
`limits. The major metapeutic goals are relief
`from pain and prevention of premature os—
`‘teoarthrosis.
`
`Using the operative procedure reported by
`'l‘onnis20 and Tennis and coworkers,2i the ac»
`
`etabulurn has to be mobilized by three os—
`teotomies, which have to be pert-buried in a.
`defined position and orientation in relation. to
`the acetabulum. If the distance from the so
`etabulum is too short there is, a higher risk of
`avascular necrosis, and, if the acetabular frag
`.ment becomes too large its free rotation may
`be impeded]?l Moreover, the inclination of
`the out planes influences the ability to rotate
`the fragment freely :in' a specific direction.
`Conventionally, the three osteotomies have to,
`be pertonned under repeated radiographic
`control. The iSChial osteotomy has to meet
`the obturator foramen and preserve the sciatic
`nerve and the sacral ligaments. Although the
`public osteOtomy is less oxides}, it is directly
`adjacent tothe femoral vein. For the manipw
`lation and repositioning of the acetabnlum‘, a
`Schanz screw has to be fixed on the acetabu-
`iar fragment. The final iliac osteotomy has to
`meet the distal part of the incisura ischiadic‘a.
`This is the longest and most difficult os-
`teotomy and has the greatest, influence onthe
`mobiiity of the .acetabular fragment and the
`initial stability of the fixation.
`For the repositioning of the acetabulum,
`empiric standard values still are based on bi~
`planar xray projections» in the antsroposterior
`(A?) and fans profile planes. According to
`Tennis at lel lateral center edge and anterior
`center edge angles of approximately 30° to
`3.5" should be achieved; Additionally, three—-
`dimensional CT imaging is recommended
`for diagnosis
`and surgical planning to
`achieve more accurate and reliable {1338532
`merit of the individual anatomic and Home
`
`chemical conditions in the weight'oearing‘
`area of the hipJ'ZtB-sm Nevertheless, the au-
`thors know of no Validated. standards based
`on three~dimensional imaging, comparable
`with the empiric data provided by, Tonnis et
`
`WMT 1006-5
`
`

`

`.
`Number 354
`
`September. 1998
`Surgery With image Based individual Templates
`33
`
`
`
`Fig“ 3. Open door: decompression in-ihe cervical spine: (A) daslgn of an lndiwduai templata (2} with
`a cap’ying pmfile (6) limiting cutting depth 0f the milling head {7) is the anterior cortical layer of the
`lamina (1); a micmmiliin‘g machine (4) is equipped with a-s‘iiding jig platé (8); (B) planning of the
`milling work on bone, arrows. indicating the specified are of bone to be removed; ((1) the; reference to
`the bone structure is provided by .an individual contact lace (9); the two individual parts of the tem-
`plate are cemented by a simple adapter (10): (D) intraoperative eXecutian in the framework of a ca~
`daver study; (E) result on the left side showing the intact-slum; and (F) result on the'righl side with the
`intact cOrllcal hinge.
`'
`
`WMT 1 005-5.
`
`WMT 1006-6
`
`

`

`
`
`34 Radermaoher et at
`
`Clinical Orthopaedics
`andReleted Research
`
`al?‘ Therefore, in addition to the conven-
`tional technique, the authors use a combina—
`tion of anterior center edge and lateral center
`edge angle measurements in AP and faux
`profile xray projections reconstructed from
`CT data. automatic measurements of the an-
`terior center edge and lateral center edge an-
`gles on the reconstructed threevdimension‘al
`hone model, and an analysis of the surface
`coverage of the femoral head in the weight~
`bearing zone. Comparison and evaluation of
`these different assessment methods are sub~
`jects of ongoing research.
`Computed tomographic scans» of the pa-
`tients normally are done at 3‘ to 4—mm slice
`distances with 30 to 40 images including one
`AP topogram. The data are transmitted to the
`DISCS planning. system with which a non—
`technieal user can generate individual tem~
`plates antenomously. The entire computer
`based surgical planning session of approxiv
`mately 5 to it) minutes is subdivided into
`sections,
`for
`three~dimensioiial diagnosis,
`planning of osteOtomi-es, Simulation: and bio!
`mechanical analysis of the repositioning, and
`the specification of appropriate reference
`Contact faces of the template on bone.u At
`the beginning'of the planning session the
`surgeon selects the type of intervention and
`the specific surgical technique. After this se-
`lection. he or- she is guided along the subse—
`quent steps of surgical planning according to
`established guidelines and handbooks
`To plan an osteotomy, the surgeon defines
`a margin of safety of 14 to 20 mm of all cuts
`from the acetabulnm. The system automativ
`Cally generates a corresponding periacetabu-
`lar safety sphere and automatically constrains
`the Specification of all three osteotornies to
`be tangential to this sphere (Fig 4A}. Thus,
`the interactive manipulation and definition of
`the osteotomiesis facilitated significantly be—
`cause ~only the inclination of the cut planes
`has to be adjusted interactively in the Sagittal
`and transverse planes, and optimal mobility
`of the acetabular fragment is. guaranteed. A?»
`ter.
`the virtual
`implementation of V the» os«
`teo‘tornies and automatic segmentation of the
`
`acetabulum and the femoral bone, the acetab~
`ular fragment Can be manipulated directly by
`the surgeon in the transverse and sagittal
`planes just as it Would be done during
`surgery (Fig 4B). For each position, the sys-
`tem automatically calculates the resulting an~
`terlor center edge and lateral center edge aria
`gles from the three-dimensional} model and
`enables three-dimensional analysis of the
`coverage of'the. femoral head; Moreover, the
`mobility of the acetabular fragment can be
`analyzed to become "aware of potential collie
`alone with the surrounding bone struatures
`and the possibilities of fixating the fragment
`canbe assessed The entire planning session
`is documented by the system, and printouts
`of all relevant planning stages and parame-
`ters, including the exact angles o'f'the’ correc‘
`tion- in the sagittal and transverse planes. nec»
`essary to achieve the planned lateral center
`edge and anterior center edge angles, can he
`provided automatically
`the Surgeon cle—
`After this, planning step,
`eides for which osteotomies he or she wants to
`use a customized {tool guide, The pubic out is
`not critical and does not require an individual
`template. Clinical trials also have shown that
`the ischial osteotomy is facilitated sufficiently
`by the preoperative planning step and the in»
`traoperative identification of the appropriate
`anatomic landmarksNevertheless, individual
`templates for the iscltial osteotomy can be
`manufactured if. desired and have been used in
`some clinical cases.“ In contrast, an individ—
`ual template always is, used for the most criti-
`cal iliac deteotomy. As the positioning of the
`template already is defined partially after plan-
`ning of the osteotomy,
`the surgeon can be
`guided by the system in an efficient and into—
`itive way. He or she uSes the three—"dimen—
`sional View to specify the direction of his or
`her surgical approach. and the desired position
`of the template on the bone The final step in-
`volves the automatic milling of the template
`Then, after 5 to 10 minutes ofinteractive sur-
`gical planning and 10 to 20 minutes. automatic
`desktop customization. (Fig. 4C), the individual
`templates can be labeled, packed, and Sterilr
`
`WMT’ 1 006-7
`
`WMT 1006-7
`
`

`

`i
`
`Number 854
`September. 1998
`
`
`
`Surgery With image Baaed Individual Templates 35
`
`m
`
`Fig 4‘, Triple pelvic osteotomy: (A) planning of the iiium asteoiomy; (B) simulation of the repositia’n—
`ing; (0) DiSQS pianning station with a desktoga maiwfacturing unit; (0) the intracperaflve imiset with
`different individual templates (-1). adaptabie stainless 'steei too! guides (2). and optionai bane pins
`'(3); (E) intraoperatiVe 'use;tand {F} radiographic wntmi; PC, a: personal computer.
`
`izsd by conventional swam, Because the tem~
`plates can be autflciaved to 135°C, they can‘be
`ready for surgery within less than 1. hour afier
`the transmission ofvthe CT image data.
`Data an 24 triple osteommies collected be~
`tween March 1996 and March 1998 are
`shown in Tabla 1. in this group, 11 intarvew
`'tions have been done in the conventional
`manner and 13 interventions have, been done
`using individual templates, In four cases mi:
`templates could not be‘positioned exactly. Be-
`cause. of very strong soft. tissue:g the comact
`
`face bstwaen the template and bone ihat had
`been specified during pianning the i11t¢rven~
`tion was not accessible without additional’ra‘
`section of 39ft tissue. This prbblem has man
`overcome by a new éesign of the templates
`and the tool guide‘, Positioning of the refer,—
`ence contact face of the; iliac. tempiate on the
`anterior supcarior iliac spine medial and cm»
`nial to the cutpl'ane allows the best access to
`the bone. surface through 31658 invasive surgi~
`cal approach. After these modifications the
`templates could be. positioned appropriately.
`
`WMT 1006-8
`
`WMT 1006-8
`
`

`

`36
`
`Radermacher at al
`
`Clinical Orthopaedics
`and Related Research
`
`TABLE 1 . Triple Osteotomy Interventions March 1996 to March 1998
`
`Number
`of
`Interventions
`
`Number of
`Interventions
`Without individual
`Templates
`
`I Number of
`intervemions
`With, individual
`Templates
`
`Optimal,
`Positioning
`of Template
`and Cut‘!
`
`Failed
`Positioning
`oi
`Templates
`
`10
`10
`4
`24
`
`7
`3
`1
`l 1
`
`8
`1 t 6*”
`31111
`'1-3»
`
`1
`5
`3
`9
`
`‘2
`2
`01:
`4
`
`Year
`
`1996
`1997
`1998
`Total
`
`‘Subiectively assessed by means of radiographs.
`“With the new integrated DiSOS'planning'and manufacmring's‘ystem,
`
`Intraoperatively the main benefit of pre-
`operative planningof individual templates is
`avoiding ,an iterative search of the optimal,
`cut planes and correction angles. under rea
`\peate'd radiographio control, better mobility
`of'the ace‘tabulumt, and a shorter time of in~
`tervention. The template can be fixed option»
`ally by a bone pin orheldby the assistant ns~
`ing- a .flieitiblo handle (Fig 4D—E). A single
`radiograph is used to confirm the po‘sitionof
`the tool guide, after which the osteotorny can
`be performed with an oscillating saw and 013+
`tion‘ally finishedrwith a chisel (Fig 4F).
`In addition to redesigning templates for a
`more medial approach to the iliac wing; nu«
`merous other changes have been made in, the
`course of developing the approach described
`in this gape; These include modified patient
`positioning (lateral and supine as described
`by Tonnis. at £11“) and the ass of a new type
`of saw (jigsaw). Only data from interwar)»
`lions that were done under comparable eon—
`dltions have been considered for additional
`analysis; All
`templates were planncd and
`manufactured using the new integrated
`DISGS station, and all patients have been
`operated on by the same surgeon, who had
`more than 12 years-eXpetience with this type
`of intervention.
`
`In a first analysis of data from 10 intern
`venti‘ons of. this type collected during an on
`going. comparative study between January
`1997 and March 1-998-(Table ,2), the mean ef~
`festive time of ,intraoperative xray monitor»
`ing was reduced by morethan.75%, and the
`
`mean time of intervention was reduced by
`25% compared with the conventional proce-
`dures without individual templates. The ex—
`act positioning of the osteotomies resulted in
`optimal mobility of the acetabulum, which
`could be repositioned according to the.
`planned values. No additional soft tissue had
`mine removed and the periosteum was pro»
`served in every case.
`
`BISCUSSION
`
`Although the number of cases has been, too
`small to derive reliable statistical data,
`the
`preliminary results are promising. The indie
`vidual template approach represents a rela-
`tively simple, low .cost; and easy to use. solu—
`tion.
`that
`facilitates precise. preoperative
`planning and. appropriate intraoperative im~
`‘plementation. The eoHVentional, intracpexa
`tive procedure is preserved and no additional
`intraoperative registration steps, computer-.
`ized equipment,
`space, or personnel are
`needed, Continual fluoroscopic monitoring
`can be avoidori, fewer radiographs are needed;
`and the duration of intervention is reduced ..
`The position of best‘fit for positiouing the
`template can be found easily by hand, be—
`cause no significant free motion of the tem—
`plate should occur while pressing the tem—
`plate slightly against thel'bione. Because of
`the structured surface 0f the templates’ con».
`tact faces and the redundancy of the surface
`based matching; the templates are suffi-
`ciently robust against. preparation artifacts
`
`-WMT 1006-9
`
`WMT 1006-9
`
`

`

`Number 354
`September. 1998,
`
`
`Surgery With Image Based IndividualTemplmes
`37
`
`TABLE 2 Preliminary Results of the Comparative Clinical Evaluation
`10 InterventionsWith StandardizedTechnique
`and Boundary Condlfle‘ns
`
`Five Interventions
`Fina Interventions With
`
` January 1 997 to March 1998 Individual Templates Without- Individuai Temp/ates
`
`
`105485
`105430.
`Time of intervention (minutes)
`157
`118
`Average (minutes)
`170
`120
`Median (minuies)
`0.08% .45
`0,064.34
`'lntIaoperative radiograph time (minutes)
`0.65
`0.15.
`Average (minutes)
`0:07
`0.23
`'MedIan (minutes)
`
`(such as small quantities of remaining soft
`*tisSue or small defects in the modeled sur~
`face} The periostcum can, be preserved.
`one main drawback of the, approach is
`that it depends. on preoperative CT imaging.
`Moreover, no percumneous applications are
`possible, (except in dental or maxillofacial
`surgery}. Lamellar contactvfa‘ce‘s (with ribs or
`arrays of pins) could. be manufactured to
`Campgns‘ate' to a limited. extent for: remaining
`soft tissue.” But as in mast sensor based sun
`face registration techniquesr the accuracy
`and reliability of this {echniqun depn'nds 0n
`{the .ac‘cassibiiity and apprnpziate intraoperaw
`tine identification of the rigid. reference
`structures. segmented am} modeled in the
`preoperative image data.
`Fumra research wiil be dedicated to the:
`ongoing clinical evaluatinn oi? the nemom
`Skater system ’er pelvic surgery in crisper»
`mien with additional clinical centers. The >
`bannfit for less expnrienced surgwns and
`the infinénce on the intervention specific
`learning curve will be investigated in more
`detail. Because postoperative 'CT scanning
`of patients docs n0: seem justifiable, acidi~
`tiona'l cadaver” studies will be conducted :0
`obtain more quantitative. data'on. the, overall
`accuracy of the technique. It is planned In
`intngrate additional tool‘s into Ina system (in
`particular for hip, knee, and spine surgery),
`magnetic mannance image processing mad,—
`
`ules and enhanced, models fOr efficient bio—
`
`mechanical analysis.
`
`References
`
`(I.
`
`I. Abel ME. Sutherland DH, Wenger? DR. ‘et rah'E‘valIIa»
`Iion of CT. scans and 3.13 Iefoxma’lted images "for;
`quantitative assessment of the hip. 3 Pediatrv Onhop.
`{4:48413 3994
`,
`2,, Axum H Taned'a H i‘garashi H: Evaluation of ac«
`- elabular coverage: Three-dimensional CT imaging
`and modifiad pals/w inlet view J Pediatr Orthop‘
`113654769 1991
`3. Bauer R Kerschbaumer F, Poise! S: OnhopadIsche
`Walicflsleme. Stuttgart, ThiemeNeflag 1994,.
`‘4; Brunner R, Robb 1: Inaccuraey nf the migratinn per
`centage and cenLer~edge angle in predicting femoral
`head IiispiacementIn cerebral 13111831.} Radian Or-
`thop 5:239L2411996
`' 5. Cinquin P, Bainville E,Barbe C, at 81: Computer as:
`sisted mehical intewencinns IBEE Eng Med Rial
`14:254~»263, 1995‘
`'Facaoaru C Fricling B; Ve‘xfahren zur Ermitllungm~
`fonnatonscher Belasmngw Z Arbeitswismnsnhafi
`39:63—72,1985
`7) Jerosch 3, Malms 3 Castro WHM Wagner R,
`Wicsncr L: Lagekomrnlle. vnn PedikelschI-auben
`flashmmumzntiermr domain: Fusion der Lendenv
`wirbalsfiulnlmthop1302479483 1992
`‘8'. Kiwi: KL, Wallin A, Ganz R: (LT eValuation of seven
`age and cong’mency of the hippnor to osteoto‘my
`Clin Onhnp 232:15~25 1988
`9'. Matsen 113 FA Garbini 1L, Sidles 3A, et-al: Robotic-
`assistance: in nnhopaedic surgery: Aproof of princi-
`ple using distal femoral aahroplasty Clin thop
`296: 178486, 1993.
`10. ,Millis MB Murphy SB: Use of computed tonic»
`graphic rcwnstmmion in planning asteotomies of
`vthehipClinOrt’hop 2741544159 1992
`l}. Ponheine F, Radermachcr K Ztmolong A et a1:
`Development of a Clinical Dcmonsnator fer
`
`WMT 1 006-1 0
`
`WMT 1006-10
`
`

`

`38
`
`Clinical Orthopaedics
`Radermaoher et al
`and Related Research
`
`Computer Assisted Orthopedic Surgery with CT
`Image Based Individual Templates. In ’Lemke EU.
`Vannier-MW, Inam‘ura K (eds). Computer As~
`vsisred Radiology and Surgery. Amsterdam, Else
`vier'944—949, 1997.
`,
`,
`,
`Radermachor K, Prothoine‘ F, Zimolong A. et al:
`Image Guided Orthopedic Surgery Using Individ~
`ual Templates~~Experimental Results and As-
`pects of the Development of a. Demonstrator for
`Pelvis Surgery.