COMPUTER-ASSISTED TOTAL KNEE
`
`REPLACEMENT ARTHROPLASTY
`
`8. DAVID STULBERG, MD, FREDERIC PICAHD, MD,
`and DOMINIQUE SARAGAGLIA, MD
`
`The reliability of techniques to position a total knee replacement (TKR) is still limited by the relative inaccuracy of the
`instrumentation. The main obstacle encountered by mechanical instrumentation systems is the inconsistency of the
`reference points. These reference points are the centers of joint articulation that will allow the establishment of the
`mechanical axis for the lower limb. These references guide the placement of the bone-cutting guides. At present, it is
`impossible to accurately locate these articular centers preoperatively. This handicaps the accuracy of the mechanical
`instruments and limits their accuracy. The goal of the total knee instrumentation procedure is to achieve cuts that are
`perpendicular to the mechanical axes of the femur and the tibia. The longevity of total knee arthroplasty is closely
`related to its intraoperative positioning. The computer~assisted procedure offers an effective and novel positioning
`method that improves the accuracy of the surgical technique of the TKR. We have chosen to present the steps of the
`computer-assisted TKR technique next to the corresponding steps of a currently available, mechanically based
`technique that is representative of many that are presently in use.
`‘
`KEY WORDS: TKFt, surgical technique, computer-assisted surgery
`
`The success of total knee replacement (TKR) surgery
`depends on several factors, including proper patient selec—
`tion, appropriate implant design, correct surgical tech
`nique, and effective perioperative care. The outcome of
`TKR surgery is particularly sensitive to variations in
`surgical technique.“ Incorrect positioning or orientation of
`implants and improper alignment of the limb can lead to
`accelerated implant wear and loosening and suboptimal
`functional performance. A number of studies have sug—
`gested that alignment errors of greater than 3° are associ—
`ated with more rapid failure and less satisfactory func-
`tional results of total knee arthroplasties?“
`Mechanical alignment guides have improved the accu—
`racy with which implants can be inserted. Although
`mechanical alignment systems are continually being re—
`fined, errors in implant and limb alignment continue to
`occur. It has been estimated that errors in tibial and femoral
`
`alignment of more than 3° occur in at least 10% of total
`knee arthroplasties, even when performed by experienced
`surgeons using mechanical alignment systems of modern
`design."2 Mechanical alignment systems have fundamen—
`tal limitations that
`limit
`their ultimate accuracy. The
`accuracy of preoperative planning is limited by the errors
`inherent to standard radiographs. With standard instrumen-
`tation, the correct location of crucial alignment landmarks
`(eg, the center of the femoral head, the center of the ankle)
`is limited during the performance of a TKR. Moreover,
`
`
`
`From the Section of Joint Reconstruction and Implant Surgery, Northwest-
`ern University. Chicago, IL: and the University Department of Orthopaedic
`Surgery, Grenoble. France.
`Address reprint requests to S. David Stulberg, MD, Professor Clinical
`Orthopaedic Surgery, Director Section of Joint Reconstruction and Implant
`Surgery, 680 N. Lake Shore Drive, Suite 1206 A, Chicago, IL 60611.
`Copyright © 2000 by W.B. Saunders Company
`1048-6666/00/1001-0005$10.00/0
`
`mechanical alignment and sizing devices presume a stan-
`dardized bone geometry that may not apply to a specific
`patient. Even the most elaborate mechanical instrumentaw
`tion systems rely on visual
`inspection to confirm the
`accuracy of the limb and implant alignment.
`Computer—based alignment systems have been devel-
`oped to address the problems inherent in mechanical total
`knee instrumentation. Although a number of computer—
`assisted TKR approaches are currently being devel~
`oped,'3"7 we have chosen to describe in detail a technique
`that
`
`O Incorporates the use of a currently available and clini-
`cally validated state—of—the-art mechanical instrumenta-
`tion system
`0 Uses commonly available, relatively inexpensive com—
`puter equipment
`(eg, a desktop computer,
`low—end
`optical localizer)
`0 Is currently available for clinical use
`0 Has available preliminary multicenter results compar-
`ing the use of the system with a mechanical system.
`These results confirm that the system is safe, that limb
`and implant alignment is superior to that achieved with
`the mechanical system, and that
`initial
`function is
`equivalent to that obtained with the mechanical sys—
`temini
`
`We have chosen to present the steps of the computer’
`assisted TKR technique next to the corresponding steps of
`a currently available, mechanically based technique, repre~
`sentative of many that are presently in use. Because a
`computenbased surgical
`technique introduces concepts
`and equipment not currently familiar to surgeons who
`perform TKR surgery, we hope that the juxtaposition of the
`2 techniques will allow surgeons to more readily under-
`stand the rationale and the function of these new tools.
`
`Moreover, by juxtaposing the 2 techniques, we wanted to
`
`Operative Techniques in Orthopaedics. Vol 10, No 1 (January). 2000: pp 25—39
`
`25
`
`rrrrrrr WMT 1005-1
`
`WMT 1005-1
`
`

`

`Fig 1. Positioning of the patient for surgery.
`
`
`
`incision and exposure.
`
`Fig 2.
`
`Fig 3., Computer system set up: localizevr, laptop or desktop, tootpedai control.
`
`26
`
`STULBERG ET AL
`
`’WMT 1005-2
`
`WMT 1005-2
`
`

`

`PATIENT POSITIONING AND SURGICAL
`EXPOSURE
`
`The mechanical alignment and computervassisted surgical
`techniques use similar approaches for patient positioning
`and surgical exposure (Fig 1). Leg holders and pneumatic
`tourniquets, routinely used with mechanical instrumenta—
`tion, can also be used for the computer-assisted technique.
`The computer—assisted technique requires that the ipsilat-
`eral iliac crest be sterilely washed and draped to allow
`placement of a screw to hold one of the rigid bodies.
`No alterations in the surgical incision usually used for
`TKR surgery ‘need to be made for the computer-assisted
`technique. Although this procedure will require the place—
`ment of rigid body holding screws in the proximal tibia
`and distal femur, the sites for these screws can be reached
`through a conventional incision and exposure (Fig 2).
`We prefer a straight midline skin incision and a medial
`para patellar exposure of the knee. This exposure extends
`distally along the medial-most edge of the quadriceps
`tendon and patella to a point just medial and distal to the
`patellar tendon insertion on the tibial tubercle. The superfi—
`cial and deep medial collateral ligament is elevated around
`the anterior medial half of the tibia, and the infrapatellar
`fat pad. The ligamentum mucosum and anterior lateral
`capsule are elevated from the anterior-lateral surface of the
`tibia. The patellar is everted laterally, and the knee is
`placed in 90° of flexion. The anterior cruciate ligament is
`resected,
`the osteophytes are removed, and the fat pad
`trimmed to allow adequate exposure of the tibia.
`
`LOCATING THE CENTERS OF THE HIP, KNEE,
`AND ANKLE JOINTS
`
`The mechanical surgical technique uses jigs and alignment
`rods to locate the centers of the hip, knee, and ankle joints.
`These determinations are made during the surgical proce~
`dure. The computer-assisted technique requires that the
`centers of these joints be determined before the positioning
`of the rods and jigs. To determine these joint centers,
`equipment unique to computer-assisted surgery must be
`used. This equipment is also used to guide the positioning
`of the cutting blocks during knee replacement.
`The equipment includes: (1) an optical localizer; (2) rigid
`bodies containing diodes; (3) 3.5—mm stainless steel bicorti—
`cal screws specifically designed to hold one of the rigid
`bodies on the bone; (4) a metal plate to hold a rigid body to
`the foot; and (5) a computer, a monitor, and a foot control.
`The localizer consists of cameras that detect the infrared
`
`radiation emitted by the diodes contained in the rigid
`bodies (Fig 3). The rigid bodies are securely affixed to the
`bones by using the bicortical screws so that they do not
`move relative to the bones when the leg is flexed, extended,
`and rotated. The localizer is connected to the computer and
`the monitor. The position of the leg and bones can be seen
`on the computer screen when the surgeon activates the foot
`control. To correctly determine the centers of the hip, knee,
`and ankle joints, it is necessary to place rigid bodies on the
`pelvis, femur, and tibia. The presence of the rigid body on
`the pelvis assures that any pelvic motion that occurs when
`the leg is moved is monitored. The position of the leg
`relative to the pelvis, therefore, is always known. When
`this combination of rigid bodies is used, it is possible for
`
`27
`
`WM-T 1005-3
`
`Fig 4. Computer system in the operating room.
`
`emphasize that the alignment objectives of the 2 tech—
`niques are identical. The goal of the computer—assisted
`system is to increase the accuracy and reproducibility with
`which the objectives of a mechanical alignment system are
`achieved
`
`
`
`SURGICAL TECHNIQUE
`PREOPERATIVE PLANNING
`
`
`
`
`
`Although the 2 techniques have very similar approaches to
`preoperative planning,
`they do differ in one significant
`respect; Both approaches require that the surgeon deter-
`
`mine the desired anatomic alignment (femoral—tibial angle)
`
`on a full—length (including hip and ankle joint), standing
`
`anterior—posterior (AP) radiograph. Some surgeons may
`also wish to determine the desired posterior slope of the
`
`tibial cut by using a lateral radiograph. This measurement
`can be used during either procedure. Many surgeons also
`
`find it helpful to estimate the desired size of the femoral
`
`and tibial implants by holding scaled templates of these
`
`implants against AP and lateral radiographs of the knee.
`The computer-assisted technique eliminates the need for
`
`this preoperative step because measurements are made
`during the procedure to determine the most appropriate
`
`implant size.
`
`COMPUTER~ASSISTED TKR
`
`WMT 1005-3
`
`

`

`
`
`Fig 5. Rotational technique to
`determine the center of the
`femoral head.
`
`‘
`
`the localizer to determine the center of the joints to within 1
`mm (Fig 4).
`The screws that hold the rigid bodies are inserted at the
`beginning of the surgical procedure. The pelvic screw is
`placed in the ipsilateral iliac crest through a small stab
`incision. The femoral and tibial screws are placed immedi—
`ately after making» the skin incision and exposing the knee
`joint. The femoral screw is inserted into the medial cortex
`just proximal to the medial femoral condyle at a 45° angle
`to the long frontal axis. The tibial screw is inserted into the
`anterior—medial cortex approximately 5 mm below the
`tibial plateau. The heads of these screws have been spe-
`cially designed to hold the rigid bodies.
`The center of the femoral head is determined by securing
`rigid bodies to the pelvic and femoral screws. The femur is
`then flexed, extended, abducted, adducted, and rotated.
`This movement generates a cloud of points on a sphere.
`The center of the sphere (ie, the femoral head) that created
`this array of points can then be imputed (Fig 5).
`The center of the knee joint is determined by securing
`rigid bodies to the femur and tibia (Fig 6). The knee joint is
`then flexed and extended and internally and externally
`rotated in 90° of flexion. This movement allows the center
`
`of rotation of the knee joint to be calculated (Fig 7). A
`surface registration technique is then used to confirm the
`center of knee joint rotation. The surgeon selects a series of
`points with the registration probe on the posterior medial
`and lateral femoral surfaces, and the anterior distal femo-
`
`28
`
`ral cortex. The rotational center of the distal femur can then
`
`be calculated and compared with the location determined
`by flexing, extending, and rotating the tibia on the femur.
`The surface registration step also provides the surgeon
`with information to be used later in the procedure. The
`optimal size of the femoral component, which is the size
`that most closely corresponds to the AP dimensions of the
`femur as measured by paipating the posterior surfaces of
`the femoral condyles and the anterior femoral cortex, is
`automatically calculated. The frontal plane of the femur
`can also be defined“ when the location and orientation of
`the posterior femoral condylar surfaces are determined.
`The size of the femoral component and the frontal plane of
`the femur are stored in the computer and provided to the
`surgeon during the preparation of the distal femur.
`The center of the ankle joint is determined by attaching a
`metal plate and an elastic band to the sole of the foot. A
`rigid body is attached to this plate. A second rigid body is
`placed in the tibial screw. The ankle joint is then flexed and
`extended. This movement allows the center of rotation of
`
`the ankle joint to be calculated (Fig 8). A surface registra-
`tion technique is used to confirm this center of ankle
`rotation. The middle of the medial and lateral malleoli and
`
`the center of the talus are palpated with the registration
`probe. These points allow the center of the ankle joint to be
`calculated. This center is compared with the location
`determined by flexing and extending the joint (Fig 9). The
`surface registration of the ankle also provides the surgeon
`
`STULBERG ET AL
`
`WMT 1005-4
`
`WMT 1005-4
`
`

`

`
`
`Fig 6. Rigid bodies secured to the femur and tibia.
`
`with information that will be used later in the procedure.
`The location of the midpoints of the malleoli and the center
`of the ankle joint allow the sagittal and frontal planes of the
`tibia to be calculated. This information is stored and
`provided to the surgeon during the preparation of the
`proximal tibia. Once the centers of the hip, knee and, ankle
`joints are determined, the femur and tibia can be accurately
`prepared.
`
`PREPARATlON OF THE TlBlA AND FEMUR
`
`The sequence of preparation is the same for the mechanical
`and computer—assisted techniques.
`
`Tibial Preparation
`
`Mechanical technique. An intramedullary or extramedul—
`lary alignment device is placed into or on the tibia. The
`medial—lateral midpoint of the tibial plateau is determined.
`The center of the tibial spine is often selected to represent
`this point. The proximal portion of the tibial instrumenta—
`
`COMPUTER~ASSISTED TKFl
`
`tion system is placed in such a way that the intramedullary
`or extramedullary rod intersects the proximal medial—
`lateral midpoint of the tibial plateau. An ankle clamp helps
`position the distal portion of an extramedullary rod over
`the midportion of the talus. This point represents the
`medial-lateral midpoint of the ankle joint. An intramedul-
`lary rod completely inserted into the tibia also rests over
`the midpoint of the talus (Fig 10).
`the
`If an extramedullary alignment system is used,
`device is placed in a position that brings the rod parallel in
`the sagittal plane to the long axis of the tibial shaft (Fig 11).
`If an intramedullary alignment device is used, the com—
`pletely inserted rod will be parallel to the long axis of the
`tibia in the sagittal plane.
`A proximal tibial cutting block attached to either the
`intramedullary or extramedullary alignment rod is placed
`along the anterior surface of the tibia proximal to the tibial
`tubercle. A stylus attached to the cutting block allows the
`surgeon to'determine the desired level of the tibial resec~v
`tion. The posterior slope of the proximal
`tibial cut
`is
`determined by selecting a block with or without a predeter—
`mined posterior slope (eg, . °).
`The rotation of the proximal tibial cut is made by placing
`the entire tibial device (rod and cutting block) so that it is
`directed along the AP line of the spine of the tibial plateau
`or, alternatively, so that it is directed toward the medial one
`third of the tibial tubercle. The cutting block is secured to
`the tibia with pins. The proximal tibial cut is then made.
`
`Computer-assisted technique. The tibial alignment guide
`used in the computer—assisted technique is virtually identi—
`cal to that used in the extramedullary mechanical align—
`ment technique described previously. This guide is posi—
`tioned against
`the leg and secured at
`the ankle with
`clamps and at the knee with 2 threaded pins inserted
`through holes on the cutting block (Fig 12). Once the
`alignment device is positioned, the computer technique is
`used to determine the orientation of the frontal and sagittal
`cuts' and the depth of the resection. Rigid bodies are
`attached to the screw in the proximal
`tibia and to the
`cutting block. The position of the cutting block relative to
`the tibia can, therefore, be calculated and depicted on the
`monitor.
`.
`
`The location of the cutting block on the tibia is depicted
`on the screen of the monitor as a red line. A green line
`represents the desired orientation of the cutting block, (ie,
`perpendicular to the mechanical axes in the frontal and
`sagittal planes). A series of wheels on the alignment guide
`allow the surgeon to tilt the cutting block until the red and
`green lines overlap (Fig 13). The block is then locked to the
`alignment guide in this position. The surgeon uses a probe
`to palpate and record the level of the medial or lateral
`plateau. The surgeon uses this information to determine
`the level of the tibial resection (eg, 2 mm below the most
`involved side). The cutting block can then be positioned,
`by using the wheels and the graphical depiction on the
`monitor, at this desired level. Once the position of the
`cutting block is determined, the device is fixed to the
`proximal tibia with 4 threaded wires. A standard tibial cut
`is made with an oscillating saw (Fig 14).
`
`29
`
`..... WMT 1005-5
`
`WMT 1005-5
`
`

`

`
`
`Fig 7. Rotational technique to
`determine the rotational cen-
`ter of the knee.
`'
`
`Femoral Preparation
`
`Distal femoral cut. Mechanical technique. An intramedul-
`lary rod is introduced into the femoral canal. The distal
`femoral cutting block device is placed onto the rod with the
`desired femoral-tibial angle (determined during the preop-
`erative planning). The device is placed against the distal
`femur, and the distal femoral cutting block is secured to the
`femur with pins (Fig 15). The cutting device determines the
`depth of the distal femoral cut (ie, distance of cut from
`distal end ofthe femur). If a surgeon wiShes to resect more
`(or less) of the distal femur,
`the cutting block can be
`repositioned on the pins.
`
`Distal femoral out. Computer-assisted technique. The femo-
`ral cutting guide must be placed in the center of the distal
`femur at the level of the knee joint before the orientation
`and depth of the femoral resections can be determined.
`Rigid bodies are attached to the screw on the distal femur
`and to the distal femoral cutting jig. The surgeon can then
`track on the monitor the position of the cutting block
`relative to the distal femur. Once the jig is centered over the
`distal
`femur,
`it
`is secured to the femur with a 5-mm
`threaded wire.
`
`The desired sagittal orientation of the femoral alignment
`device is then determined. A clashed green line within a
`square box on the monitor represents the ideal sagittal
`position of the device, which is perpendicular to the
`femoral mechanical sagittal axis. A red line in the box on
`the monitor represents the sagittal position of the device.
`The surgeon manually positions the alignment device until
`the 2 lines overlap. When this device is correctly posi—
`tioned,
`it
`is secured to the femur; This establishes the
`
`sagittal orientation of the distal femoral cutting block (Fig
`16A and B).
`Two keels are secured to the femoral alignment device.
`The distal femoral cutting block is then attached to the
`
`30
`
`device. The frontal orientation of the cutting block is then
`established. A green line within a circle on the monitor
`represents the ideal
`frontal position within the circle
`represents the cutting block. A wheel system on the
`femoral alignment system allows the surgeon to change
`the position of the cutting block. When the 2 lines overlap,
`the cutting block is then secured firmly to the alignment
`device. The desired depth of the femoral resection was
`calculated by the system when the rotational center of the
`knee joint was determined. The location of the green line
`takes this depth intoraccount. Therefore, once the 2 lines
`overlap, the correct depth of the distal femoral cut is fixed
`
`
`
`Fig 8. Rotational technique to determine the rotational center
`of the ankle.
`
`STULBERG ET AL
`
`WMT 1005-6
`
`WMT 1005-6
`
`

`

`Fig 9. Surface registration
`step.
`
`
`
`(Fig 17A and B). The distal femur is then resected. The
`distal cutting block is then removed from the femoral
`alignment device.
`
`Establishing Femoral Component Rotation
`
`Mechanical system. Most systems place the femoral com—
`ponent so that its medial~lateral aids is parallel
`to the
`epicondylar axis. If the posterior femoral condyles are not
`deformed, a jig with projections that rest against
`the
`posterior condyles of the femur can be used to establish the
`rotation of the component. The epicondylar axis is exter—
`nally rotated 3° from the line connecting the posterior
`surfaces of the medial and lateral condyles. Mechanical
`systems incorporate this relationship into "the jig with
`projections that rest against the posterior condyles. Once
`the desired position of the jig is established, the holes for
`the pegs of the AP cutting block are made with pins or a
`drill (Fig 18A and B).
`
`Computer—assisted system. The locations of the medial
`and lateral femoral epicondyles as well as the location of
`the posterior surfaces of the medial and lateral femoral
`condyles were determined during the surface registration
`
`COMPUTER-ASSISTED TKR
`
`step of the procedure. This information is now used to
`‘ establish the desired rotation of the femoral component.
`The femoral alignment device that was previously secured
`to the distal femur can be rotated by using the wheel
`system. A circle is displayed on the monitor with a green
`line representing the ideal rotational alignment of the
`femoral component and a red line representing the rotation
`of the alignment device. The wheel system is used to
`exactly overlap these 2 lines. The device is then locked in
`this position (Fig 19A).
`
`Establishing the Size of the Femoral Component
`
`Mechanical system. Most systems use referencing guides
`off the posterior femoral condyles and/or anterior femoral
`cortex to establish the size of the femoral component. A
`posterior—based referencing system uses a jig with projec—
`tions that rest against the posterior condyles of the femur.
`This jig is placed against the resected distal femur. A stylus
`placed into the anterior surface of the jig is then used to
`select the appropriate femoral component size.
`
`Computer-assisted system. The optimum size of the femo—
`ral component was calculated at the time that the rota-
`
`31
`
`WMT 1005-7
`
`WMT 1005-7
`
`

`

`the knee joint was determined. This
`tional center of
`information is now provided by the system to the surgeon
`(Fig 19A).
`
`Anterior—Posterior Femoral Resection
`
`Mechanical system. The cutting block corresponding to
`the femoral component of desired size is placed into the
`holes on the distal femur. The anterior, posterior, and
`chamfer cuts are then made (Fig 19B).
`
`Computer-assisted system. The cutting block correspond—
`ing to the femoral component of correct size is attached to
`the distal femoral alignment system. The correct rotational
`and frontal position of this device has been previously
`
`
`
`Fig 10. An intramedullary alignment device is placed into the
`tibia during the mechanical technique:
`
`32
`
`
`
`
`
`Fig 11. Extramedullary alignment system used. in the me-
`chanical technique.
`
`determined. The anterior, posterior, and chamfer cuts are
`then made.
`
`Trial Reduction
`
`Mechanical and computer-assisted systems. The tibial
`trial base plate that most accurately fits the resected
`proximal tibial surface is selected and placed on the top of
`the tibia. A polyethylene insert is placed onto the base
`plate. The femoral trial is placed on the distal femur. The
`knee is flexed and extended and stressed medially and
`laterally. Correct soft tissue balance is achieved through a
`combination of soft
`tissue releases and alterations of
`
`polyethylene thickness.
`The computer-assisted systems allow the alignment of
`the knee to be checked at this point in the procedure. The
`rigid bodies are fixed on the femur and the tibia and the
`final alignment of the limb can be determined (Fig 20).
`
`STULBERG ET AL
`
`WMT 1005-8
`
`WMT 1005-8
`
`

`

`PRELIMINARY CLINICAL RESULTS
`
`We tested this system on a series of 30 patients with
`primary osteoarthritis of the knee. 15 patients received the
`computer-assisted procedure (group 2) and fifteen patients
`received the classical procedure (group 2). The trial was a
`prospective,
`randomized parallel study performed in
`Grenoble, France from January 13, 1998, to December 1,
`1998.
`
`PATIENTS
`
`Thirty patients between 55 and 89 years of age (mean age,
`69) were included. There were 14 right knees and 16 left
`knees that were surgically treated in the study after
`checking eligibility (e.g., inclusion and exclusion criteria,
`signing the consent form).
`
`METHOD
`
`The same surgeon treated the 30 patients, using either the
`classical procedure or
`the computer-assisted system
`(OrthoPilot; Aesculap AG, Tuttlingen, Germany). Two
`independent surgeons followed-up the patients 6 weeks
`after surgery. The review criteria included radiological
`criteria (the main criterion was the femoral tibia mechani—
`cal angle on the long—leg radiograph), complications, and
`
`
`
`Fig 12. An extramedullary alignment device for the tibial cut
`used during the computer-assisted, technique. The screen
`displays the jig position relative to the knee center.
`
`
`Fig 13. The cutting block is secured after checking the
`correct position by using the graphical interface depicted on
`
`the screen. The dashed line representing ideal cut: The full
`line indicates the jig position. 1, frontal plane; 2, sagittal
`
`plane; 3, cutting height.
`»
`
`surgical criteria, such as duration of the procedure and
`postoperative bleeding.
`
`RESULTS
`
`Radiographs were evaluated to determine what percent-
`age of patients had (1) femoral~tibial angles between 3° of
`varus and 3° of valgus (group 1 = 100%, group 2 = 66.6%);
`(2) a femoral implant angle of 90° to the coronal mechanical
`axis (group 1 = 46.6%, group 2 = 0.06%); and (3) a tibia
`resection angle within 2° of varus or valgus (group
`1 = 100%, group 2 =1 86.6“").
`The average duration of the procedures was 101 minutes
`in group 1 and 74 minutes in group 2. The average blood
`loss of the .2 groups was the same. There were no complica-
`tions in group 1 and 3 complications in group 2 (2 patients
`had deep venous thrombosis, one patient had stiffness).
`
`CONCLUSION
`
`An overview of a computer—assisted TKR surgical tech—
`nique has been presented to illustrate the principles of this
`new technology. The system that has been described is now .
`in clinical use. Preliminary results with the system indicate
`that the alignment results that have been achieved are
`more accurate and reproducible than the mechanical sys-
`tem used to insert identical implants. The system is simple
`to understand and the computer hardware used is rela-
`tively inexpensive. The system uses currently available
`mechanical total knee instruments. The surgeon can use
`these instruments in the conventional way at any point in ,
`
`COMPUTER-ASSISTED TKR
`
`*
`
`33
`
`WMT 1005-9
`
`WMT 1005-9
`
`

`

`mwnmliun
`
`{mi tuning;-
`
`Fig 14. The cutting bloek can then be positioned. (A) Mechanical technique. (8) Computer-assisted technique. (C) Positioning
`of the tibial keel.
`
`34
`
`STULBERG ET AL
`
`,,,,,,,,,,,,,,,
`
`,,,,,,,,,,,,,,,, A
`
`_
`
`,_
`
`WMT 1.005- 1.0
`
`WMT 1005-10
`
`

`

`
`
`Fig 15. An intramedullary rod is introduced into the femoral
`canal, the distal, femoral cutting block device is then placed,
`and the distal cut is made in the mechanical technique.
`
`
`Fig 16. The distal femoral cutting block
`device is placed in the sagittal plane during
`i-tanumkulmm dwiriclmuml
`
`the computer-assisted technique. The
`,
`dashed line represents the ideal cut. The full
`
`line indicates jig sagittal position. 1-, 2—, 3-
`indicate steps to secure the distal femoral
`
`cutting block device. (A) Coronal view. (B)
`Sagittal view.
`
`
`
`
`
`COMPUTER~ASSISTED TKR
`
`35
`
`WMT 1005-11 ,
`
`WMT 1005-11
`
`

`

`41m
`
`{Limit
`
`ul‘dcu medium
`
`'
`
`Fig 1?, {A and 8); The diam: femoral crumng
`biock'device is placed in the {rental piane
`during ma mmpuiemssiswd iecbmque, The
`tibiai cm is than mack. The: dashed fine
`indicates was: auzfimaiufi fine indieatex the
`its; imam} amnion.
`
`Fig 18.
`
`(A and B). Sizing of the femoral implant and establishing femoral component rotation during the mechanical technique.
`
`36
`
`'
`
`STULBERG ET AL ~
`
`'
`
`M ,, WMT100,,,,, -12
`
`WMT 1005-12
`
`

`

`' Fig 19.
`
`(A) Sizing of the femoral implant and establishing
`femoral component rotation during the computer-assisted
`technique. The dashed line indicates Athe reference line. The
`full line indicates the jig frontal position. (B) Chamfer cuts.
`
`i mum Rul'
`
`COMPUTER—ASSISTED TKR
`
`37
`
`WMT 1005-13
`
`

`

`
`
`head Linc
`
`~
`
`
`
`(A) Trial reduction with the use of the computer~
`Fig 20.
`assisted surgery technique. (8) Graphical interface displayed
`simultaneously with computer~assisted surgery technique. (C)
`Trial reduction with the use of mechanical surgery technique.
`
`38
`
`.
`
`.
`
`'
`
`STULBERG ET AL
`
`WMT 1005-14
`
`

`

`the surgical procedure if the guidance provided by the
`computer—assisted devices is considered inappropriate.
`If systems such as the one described here are to be truly
`useful,
`they must be safe, accurate, easy to use, and
`cost-efficient. If these goals can be achieved, such systems
`are likely to be widely accepted by surgeons who perform
`TKR surgery.
`
`REFERENCES
`1.
`
`2.
`
`Freeman MAR, Todd RC, Bamert P, et al: lCLH—Arthroplasty of the
`knee: 1968—1977.] BoneJoint Surg 608:339’344, 1978
`lnsall JN, Binazzi R, Soudry M, et al: Total knee arthroplasty. Clin
`Orthop 192:13-22, 1985
`lnsall JN, Ranawat CS, Aglietti P, et al: A comparison of four models of
`total knee—replacement prostheses. J Bone Joint Surg Am 58:754—765,
`1976
`lnsall J, Scott WN, Ranawat CS: The total Condylar prosthesis. A
`report of the hundred cases. J Bone Joint Surg Am 61:173-179, 1979
`. Goodfellow JW, O'Connor J]: Clinical results of the Oxford knee. Clin
`Orthop 205:21—42, 1986
`Ranawat CS, Adjei OB: Survivorship analysis and results of total
`condylar knee arthroplasty. Clin Orthop 22616—13, 1988
`. Feng EL, Stulberg SD, Wixson RL: Progressive subluxation and
`polyethylene wear in total knee replacements with flat articular
`surfaces. Clin Orthop 229:60—71, 1994
`. Ecker ML, Lotke PA, Windsor RE, et al: Long-term results after Total
`Condylar knee arthroplasty. Significance of radiolucent lines. Clin
`Orthop 216:151-158, 1987
`the
`. Ritter MA, Herbst SA, Keating EM, et al: Radiolucency at
`bone~cement interface in total knee replacement.J BoneJoint Surg Am
`76:60~65, 1994
`Laskin RS: Total Condylar knee replacement in patients who have
`
`.
`
`.
`
`10.
`
`11.
`
`12.
`
`13.
`
`14.
`
`15.
`
`16.
`
`1’7.
`
`18.
`
`19‘
`
`20.
`
`'21.
`
`rheumatoid arthritis. A ten—year follow-up study. J Bone Joint Surg
`Am 72:529—535, 1990
`Jeffery RS, Morris RW, Denham RA: Coronal alignment after total
`knee replacement] Bone Joint Surg Br 73:709-714, 1991
`Matsen FA, Garbini JL, Sidles JA, et al: Robotic assistance in orthopae-
`dic surgery. (A proof of principle using distal femoral arthr0plasty).
`Clin Orthop 2962178486, 1993
`Kienzle TC, Stulberg SD, Peshkin M, et al: A computer assisted total
`knee replacement surgical system using a calibrated robot, in Taylor
`RH, et al (eds): Computer integrated Surgery. Cambridge, MA, MIT
`Press, 1995
`Fadda M, Bertelli D, Marteili S, et al: Computer assisted planning for
`total knee arthroplasty, in First Joint Conference of Computer Vision
`Virtual Reality and Robotics in Medicine and Medical Robotics and
`Computer Assisted Surgery, 1997. Grenoble, France, Springer, 1997,
`pp 619-628
`Glozman D, Shoham M, Fischer A: Efficient registration of 3—D objects
`in robotic-assisted surgery proceedings. Comput Aided Surg (in
`press), 1999 (abstr)
`Davies BL, Harriss J, Lin W], et al: Active compliance in robotic
`surgery—The use of force control as a dynamic constraint. J Eng Med
`211: H4, 1997
`Delp SL, Stulberg SD, Davies 13, et al: Computer assisted knee
`replacement. Clin Orthop 35414956, 1998
`Picarcl F, Leitner F, Raoult O, et al: Clinical evaluation of computer
`assisted total knee arthroplasty. Comput Aided Surg (in press), 1999
`(abstr)
`Picard F, Leitner F, Raoult O, et al: Early clinical results with the
`Orthopilot System. Comput Aided Surg (in press), 1999 (abstr)
`Stulberg SD, Picard F: The rationale, surgical technique a