A technical innovation for improving identification of the trackers by the LED cameras in navigation-assisted total knee arthroplasty

Abstract

Objective: To reduce the operating time in computer-assisted navigated total knee replacement (TKR), by improving communication between the infrared camera and the trackers placed on the patient.

Materials and Methods: The innovation involves placing a routinely used laser pointer on top of the camera, so that the infrared cameras focus precisely on the trackers located on the knee to be operated on. A prospective randomized study was performed involving 40 patients divided into two groups, A and B. Both groups underwent navigated TKR, but for group B patients a laser pointer was used to improve the targeting capabilities of the cameras.

Results: Without the laser pointer, the camera had to move a mean 9.2 times in order to identify the trackers. With the introduction of the laser pointer, this was reduced to 0.9 times. Accordingly, the additional mean time required without the laser pointer was 11.6 minutes.

Conclusion: Time delays are a major problem in computer-assisted surgery, and our technical suggestion can contribute towards reducing the delays associated with this particular application.

• Total knee arthroplasty
• navigation
• computer assisted surgery

Introduction

The use of Computer Assisted Orthopaedic Surgery (CAOS) in total knee replacement (TKR) attempts to address the problem of correct alignment and rotation of the components. Many studies [1–7] have demonstrated the advantages of using computer-assisted navigation techniques in knee arthroplasty, which in turn may improve the survival of the prosthesis. Nevertheless, navigation has been proven to be a time-consuming procedure, requiring familiarization with the software, collaboration of the surgeon and the assistant with the operating theater staff, and occasionally technical back-up. As a result, many surgeons avoid using computer-assisted navigation for TKR, even when they have had the opportunity to try it. They rely instead on traditional surgical skills to introduce the components of total knee prostheses properly. The present study introduces a simple method for accelerating the procedure of communication between the infrared camera system and the trackers, which will also reduce significantly the duration of the operation.

Patients and methods

The navigation system (Stryker Leibinger, Software 3.01) used in our study is a wireless and non-image-based system. It does not require any CT or intraoperative fluoroscopic images. The surgeon determines the mechanical alignment intraoperatively by localizing the centers of the hip, knee and ankle. The navigation system has three boom-mounted infrared cameras (Stryker Leibinger, Freiburg, Germany) (Figure 1A). These cameras are able to track battery-powered wireless trackers, using the instruments’ light-emitting diodes (LEDs) to communicate with the camera's localizing system (Figure 1B). The trackers are rigidly fixed to the bone, so that their position remains constant throughout the procedure. There is also the option to have one tracker remain on the bone while the other is attached to a mobile platform and used to check the orientation of the surface (e.g., the orientation of the tibial cut level). Using this special pointer, the surgeon can guide the procedure by navigating a menu on the computer screen. This method adds to the versatility and accuracy of the process, because the surgeon does not have to change his perspective, press a pedal, or ask someone else to press buttons on the computer keyboard. The positional information from the trackers, attached bone and instruments are analyzed by the computer, which determines the real-time position of the leg within a 3D coordinate system.

Figure 1. (A) The camera of the system with the laser pointer mounted on top. (B) The trackers fixed on the bone in a TKR operation. (C) The camera identifies the trackers. (D) The laser mark on the lateral femoral condyle approximately 8 cm below the surface of the articular cartilage. If the laser pointer aims in this area, the trackers will be immediately identified by the camera. [Color version available online.]

After the patient's details have been entered, the pointer and trackers are registered and the anatomical landmarks defined. In this initial registration of the trackers, an assistant holds the infrared camera boom by the handle (Figure 1A) and tries to move it until the camera recognizes both the trackers and the pointer. Unfortunately, this initial registration phase may require several changes of camera position, and the camera sometimes needs to be moved during later phases of the operation as well. At this stage a simple laser pointer, placed on top of the boom with the cameras, is introduced. This is an ordinary, portable, pen-shaped laser pointer (deep red laser diode near 670/650 nm wavelength). The pointer is fixed to the boom using simple adhesive tape. It is positioned in the center (above the middle camera) and aimed at the knee (Figure 1). One assistant, who is not scrubbed, presses the button of the laser pointer and moves the boom with the cameras until the light source from the laser pointer is focused on the medial femoral condyle of the knee. As in our cases, it is not necessary for the same assistant to participate in every case. In this position (Figure 1C), all the trackers are immediately recognized by the infrared cameras and the operation can begin with no need for further movements of the boom.

Forty patients were enrolled in this study, selected at random from candidates for TKR. All patients had osteoarthritis, and underwent a TKR using a computer-assisted navigation technique. Patients were divided into two groups: Group A was the control group, for which a navigated TKR was performed; while group B also underwent navigated TKR, but for these patients a laser pointer was used to improve the targeting capabilities of the cameras. Randomization to ascertain which patients should be assigned to which group was done in permutation blocks. All patients were operated by the same senior surgeon (K.E.) using a standardized operative technique. The surgeon had used the navigation system in more than 100 knee replacements prior to the present study. All patients received a Scorpio (Stryker) total knee prosthesis. In addition to the 40 patients used in the main trial, a pilot study with six cases was performed to ascertain precisely where the laser pointer should aim. It was discovered that the aiming point should be the medial femoral condyle, approximately 8 cm below the articular cartilage (Figure 1D).

Results

Table I shows the two groups, together with the number of times the camera had to change its position in order to identify the trackers. For group A, the average additional time following changes in camera position was 11.6 min. Each time the handle of the camera had to move, this additional time was recorded intraoperatively by the use of a chronometer. The total time required for the operation was not recorded, because only delays caused by camera transposition, as opposed to other reasons, were required for the study. The results were assessed using Fisher's Exact test and the Mann-Whitney U test. In group A, the handle of the camera had to be moved an average of 9.2 times (range: 5–12), whereas in group B the mean number of times was 0.9 (range: 0-3). Similarly, for group A the additional time was 11.6 min (range: 8–16 min); in group B, with the laser pointer, it was only 0.6 min (range: 0–2 min) (Figure 2A). Employing Fisher's Exact test, we found that in group A the camera was 1.82 (95% confidence interval 1.223–2.703) times more likely to change position than in group B (Figure 2B). In summary, group B had lower values than group A, and these results are highly significant (p < 0.0001).

Figure 2. Graphic representations of (A) the number of times the boom had to change position in both groups with the 95% confidence intervals; and (B) the number of times the boom had to change position at least once in both groups.

Table I.  Patients who underwent a total knee replacement using computer-assisted navigation with (Group B) and without (Group A) the use of a laser pointer. (Patients in each group were numbered after randomization.)

Group A (without laser pointer)			Group B (with laser pointer)
Patient No.	Times boom changed position	Additional time (min)	Patient. No.	Times boom changed position	Additional time (min)
1	9	12	21	2	1
2	11	14	22	2	1.5
3	9	8	23	1	0.5
4	10	13	24	0	0
5	12	16	25	0	0
6	5	8	26	0	0
7	11	16	27	1	1
8	8	10	28	0	0
9	9	10	29	0	0
10	9	8	30	0	0
11	11	15	31	1	0.50
12	10	14	32	2	1.50
13	8	9	33	0	0
14	9	11	34	0	0
15	7	10	35	3	2
16	12	15	36	1	0.50
17	6	8	37	0	0
18	9	12	38	2	1
19	11	13	39	1	1.50
20	7	9	40	1	1

Discussion

It has been asserted by Sikorski and Chauhan that computer-assisted navigation has the potential to provide a much higher degree of accuracy and control than is otherwise possible [1]. Subsequently, the long-term results of TKR depend on the accuracy of component placement and limb alignment [2–8]. Unfortunately, innovative technology and familiarization with computerized techniques usually requires more time initially, which can lead to its rejection by some surgeons. In an effort to simplify this technique, we introduced our innovation, which has had a significant impact on overall operating times.

Most of the pioneers of computer-assisted navigation in TKR noted an increase in total operating times. Haaker et al. [3] reported a mean increase of 10 min, and predicted that this delay would increase in the near future, when all intraoperative stages of knee replacement (femoral component rotation, AP positioning, soft tissue tension in flexion and extension) would be navigated. Anderson et al. [9] calculated that navigation increased tourniquet time by approximately 15 min. Kim et al. [10] reported that the initial phase of the anatomic survey (exactly the phase that introduction of the laser pointer was intended to improve) added approximately 15–20 min. Victor et al. [8] also compared computer-assisted navigation in TKR with non-navigated groups. They concluded that overall tourniquet time in the navigation group was 12 min longer, and skin-to-skin time 10 min longer, than the non-navigation group. The current innovation has managed to reduce the operating time by an average of approximately 11.6 min. Time-saving initiatives have a direct impact on TKR morbidity, as they affect tourniquet time, exposure of the surgical wound to the outside environment, and total blood loss (for those who do not use tourniquet). The delays may be greater if less experienced surgeons perform a navigated TKR, but, in our estimation, these cases will benefit most from the reduction in operating time achieved through this innovation.

The surgeon's experience and surgical skills will undoubtedly improve as more computer-assisted navigation procedures are performed. Nevertheless, during the initial registration phase for the trackers, it is not the surgeon who moves the handle of the camera but a nurse or assistant, and thus the time spent on this initialization process is unrelated to the surgeon's experience. In our institution, the same assistant was not used to move the handle of the boom in every case. In this crucial initial period (when most surgeons abandon the navigated technique and continue with the conventional jig-assisted technique), the addition of the laser pointer and its simplified use can offer substantial benefits, even for inexperienced operating theater staff.

Another important issue is that in all the studies with computer-assisted navigation [3], [8–10] not all stages of the operation were performed with navigation. The duration would be extended if all stages of a TKR were performed using navigation. With the simple addition of a laser pointer, it is our belief that the surgeon can reduce operating times by approximately 11.6 min. It should be noted here that the average time-difference when using the laser pointer was calculated during the navigation of all stages of a TKR procedure (femoral cut, tibial cut, soft tissue balancing, and final recording of the range of motion achieved).

Conclusions

Computer-assisted navigation has brought promising results in total knee replacement. Nevertheless, for the time being, this surgical procedure remains an innovation, and in this context all initiatives aimed at reducing the operating time can be beneficial for the treating surgeon and, of course, the patient. The initial registration phase of the trackers remains the main time-consuming portion of the whole navigated procedure, and our technical suggestion can reduce this period significantly.