Mixed-Reality Guidance for Total Knee Arthroplasty Is Associated With More Accurate Femoral Resection but Longer Operative Time Compared to Accelerometer-Based Guidance Authors: Michael A. Hewitt, Catelyn A. Woelfle, Anthony H. Zou, Wyatt H. Buchalter, Alexander L. Neuwirth, Carl L. Herndon, H. John Cooper, Nana O. Sarpong Published: 2026 ABSTRACT BackgroundMixed-reality guidance for total knee arthroplasty (MR-TKA) offers a novel approach that may enhance intraoperative planning and surgical site tracking compared to established techniques, such as accelerometer-based guidance (AB-TKA). This study investigates operative time, resection accuracy, and short-term clinical outcomes of MR-TKA compared with AB-TKA.MethodsMR-TKA procedures were performed at a single institution and retrospectively matched 1:1 with an AB-TKA cohort based on age (±5 years), sex, and body mass index (±5 kg/m2). Postoperative femoral and tibial resection angles were radiographically measured and compared with intraoperative plans. Operative times, complications, and postoperative range of motion were also recorded.ResultsA total of 94 cases were included (47 MR-TKA, 47 AB-TKA). There was no difference in age (74 ± 7.6 years; range, 53.0–88.7), body mass index (28.7 ± 4.6; range, 19.0–42.0), or sex (74.5% women, 25.5% men) between groups. Mean operative time was longer for MR-TKA compared to AB-TKA (109 ± 21 vs 90 ± 13 minutes, P < .001). Postoperative deviation from the planned femoral resection angle was lower for MR-TKA compared to AB-TKA (1.20° ± 1.21° vs 1.87° ± 1.48°, P = .021), but there was no difference in tibial resection angle (1.06° ± 0.77° vs 1.39° ± 1.08°, P = .093). There was no difference in 3-month postoperative knee flexion, knee extension, or complication rate between groups.ConclusionsMixed-reality guidance was associated with improved femoral resection accuracy but may require longer operative time compared to accelerometer-based guidance. Both techniques demonstrate comparable short-term range of motion and complication rates. Introduction Component alignment is an important aspect of total knee arthroplasty (TKA) that influences postoperative survivorship and function. There is continued debate on whether mechanical or alternate alignment strategies result in better outcomes [[1], [2], [3], [4]]. Mechanical alignment, which aims at achieving a neutral hip-knee-ankle axis, has served as a gold standard for many decades [3,5]. However, recent studies have shown that restoring physiologic knee motion through alternate alignment strategies may improve clinical results without decreasing survivorship [[6], [7], [8], [9]]. Regardless of alignment strategy, technological advances in TKA guidance systems improve a surgeon’s ability to achieve their planned alignment [10,11]. Computer-assisted guidance for TKA has emerged as a common method used to execute alternate alignment strategies over the last 3 decades [12] and allows surgeons to quantify alignment intraoperatively and assess factors such as soft tissue balance and knee morphology [8]. Computer-assisted TKA has shown benefits over traditional guidance in some studies in improving bony resection and component alignment, as well as decreased early revision rates [13,14]. However, several barriers to computer-assisted TKA adoption include additional device training and increased device cost and operative time [15,16]. Large-console, camera-based computer-assisted TKA guidance systems offer improved intraoperative guidance through the use of tracking arrays but have limitations such as requiring line-of-sight for camera function, a large spatial footprint in the operating room, and high cost [12]. A desire for smaller, more cost-effective computer-assisted TKA systems has spurred the development of other technologies such as hand-held accelerometer-based (AB-TKA) guidance systems. While accelerometer-based AB-TKA guidance is more convenient and cost-effective, most contemporary systems do not provide data on soft tissue tension or femoral component rotation [12,17]. Augmented reality is an emerging technology that has shown promise in early surgical applications by projecting images, such as virtual knee models or surgical plans, onto the operative space via a headset [18]. Recent studies in augmented-reality TKA have demonstrated excellent accuracy of bony resections [19] and alignment [20], highlighting the promising role of advanced technology in arthroplasty. Within the field of augmented reality, mixed reality refers to programs that allow direct interaction between physical and digital environments. In arthroplasty, mixed-reality TKA (MR-TKA) is a promising technology that enables intraoperative planning and tracking of optical arrays similar to large camera-tower computer-assisted guidance, but with the size and cost advantages of smaller, hand-held devices like accelerometer-based guidance since the entire system is housed within a single portable headset. MR-TKA is now available for clinical use, but limited literature is available comparing the efficacy of augmented-reality TKA or MR-TKA to other more portable guidance systems, such as AB-TKA [21]. Therefore, the purpose of this study was to investigate the operative time, resection accuracy, and early clinical outcomes after MR-TKA compared to a matched-controlled AB-TKA population. We hypothesized that MR-TKA would result in similar resection accuracy, 3-month range of motion (ROM), and complication rates compared to AB-TKA. Material and methods Study design All TKA procedures were performed at a single institution by 2 fellowship-trained adult reconstructive orthopaedic surgeons. TKA procedures were retrospectively identified and reviewed following study approval from our institutional review board (IRB number AAAV0808). A waiver of informed consent was authorized by the institutional review board, given the retrospective nature of this study. All primary TKA procedures indicated for osteoarthritis were eligible for inclusion in the study. Procedures were excluded from analysis if the patient underwent simultaneous bilateral TKA or had a prior open surgery on the ipsilateral knee of the TKA. Study population All MR-TKA procedures included in this study were consecutively performed from March 2024 to October 2024 and included the initial procedures performed with this technology for both surgeons following Food and Drug Administration clearance. After identification, MR-TKA procedures were matched to AB-TKA procedures previously performed from 2022 to 2024, with demographics controlled by age (±5 years), body mass index (BMI) (±5 kg/m2), and sex. All AB-TKA cases, including initially performed cases with a learning curve, were evaluated, and a match with the closest demographics to each MR-TKA patient was selected, thus making the AB-TKA cases nonconsecutive. The BMI criteria were expanded to ±7 kg/m2 for 1 MR-TKA patient who did not have a demographic match with the BMI ±5 kg/m2 criteria. Implant and guidance systems used All TKA procedures in both groups were performed using the Persona Total Knee System from Zimmer Biomet (Warsaw, IN). We employed a restricted kinematic alignment philosophy, with a bound of 3 degrees of varus on the tibia and 3 degrees of valgus on the femur. MR-TKA was performed using the STELLAR knee mixed-reality guidance system from PolarisAR (Miami, FL, see Fig. 1a-c). AB-TKA was performed using the KneeAlign, a portable, accelerometer-based computer-assisted guidance system from OrthAlign (Irvine, CA, see Fig. 1d). All procedures, when possible, were performed as cemented, cruciate-retaining TKAs with a medial congruent insert. One of the 2 surgeons used selective patellar resurfacing, while the other used resurfacing in all cases. No tourniquets were used. All procedures followed the same multimodal pain management protocol after surgery. Both surgeons applied the same restricted kinematic alignment philosophy to minimize soft tissue release and achieve balance primarily through bony cuts; there was no difference in soft tissue release protocol between groups.Figure 1(a) Intraoperative mixed-reality dashboard, showing planned and live alignment metrics, manipulated via hand gestures and voice. (b) Femoral and tibial-mounted optical tracking arrays used for MR-TKA. (c) Surgeon using an optical see-through, head-mounted mixed-reality display. (d) AB-TKA unit affixed to the proximal tibia (a) Intraoperative mixed-reality dashboard, showing planned and live alignment metrics, manipulated via hand gestures and voice. (b) Femoral and tibial-mounted optical tracking arrays used for MR-TKA. (c) Surgeon using an optical see-through, head-mounted mixed-reality display. (d) AB-TKA unit affixed to the proximal tibia Radiographic alignment measures All procedures were performed using a functional alignment approach with the goal of minimizing the need for soft tissue releases. The planned distal femoral resection angle and proximal tibial resection angle relative to the mechanical axis were recorded in the operative report for each patient. Follow-up radiographic alignment was evaluated using a 3 joint standing lower extremity standing radiograph within 3 months of the procedure. Deviation in distal femoral resection angle and proximal tibial resection angle was determined. All radiograph measures were determined independently by 2 investigators, and inter-rater reliability (IRR) was assessed using a calculated intraclass correlation coefficient. The average measurement between observers was used for comparison to the intraoperative plan. All radiographic measures were collected at once using a list of medical record numbers without patient information, functionally masking the patient cohort to reduce the potential for bias. However, investigators were not formally blinded, given that radiographic images were accessed directly from the electronic medical record. Operating time and clinical outcomes The operative time of each procedure was defined as the time from the first incision to final incision closure. In addition to comparing total operative time between groups, we performed a subanalysis evaluating the operative time of the first half of performed MR-TKA cases compared to the second half of MR-TKA after more experience, a method used to evaluate learning curve in prior orthopaedic literature [22]. A subanalysis comparing the first and second halves of AB-TKA cases was not performed because the matched cases were not performed consecutively. Knee ROM was measured clinically with a goniometer both preoperatively and postoperatively. ROM measurements were collected passively with the patient supine in the clinic by the attending surgeon. ROM measures were collected once by one surgeon, and IRR was not assessed. ROM measurements were not blinded given that these data were collected clinically and the retrospective nature of this study. Patient records were followed for 3 months postoperatively, and any complications requiring reoperation or revision were recorded. Power analysis Radiographic alignment was the primary study outcome and used to anchor the power analysis. In a study evaluating the same AB-TKA guidance system used in our study, Nam et al. reported an average deviation in femoral resection angle from the intended goal of 0.8° ± 0.6° [14]. Defining both the expected radiograph standard deviation and equivalence limit as 0.6, we calculated a required total population size of n = 52 (26 per group) required using an equivalence trial power calculation (applying 90% power and alpha 0.025) [23]. At the time of our study, we had more than 52 patients available and elected to include all eligible patients to increase the power of our study and avoid selection bias. Data analyses Statistical analysis was completed in the R statistical package (v. 4.2.2) [24]. A chi-square test of independence was used to evaluate all categorical variables. For comparisons of changes in continuous variables within a group (eg, change in ROM), a paired t-test was employed. For comparisons of continuous variables between groups (eg, radiographic measurements), an unpaired t-test was utilized. A P value of less than 0.05 was considered statistically significant for all analyses. Results Study population A total of 94 TKA cases were included in this study; 47 consecutive MR-TKA cases met the criteria for study inclusion and were paired with 47 AB-TKA cases based on age, sex, and BMI. The screening diagram is shown in Figure 2. Our final cohort was determined to be sufficiently powered based on our power analysis (52 cases required). There was no difference between the MR-TKA and AB-TKA groups in match-controlled age (74 ± 7.4 years [range, 54–89] vs 74 ± 7.8 years [range, 53–89], P = .99), BMI (28.6 ± 5.04 [range, 19.0–42.0] vs 28.8 ± 4.07 [range, 19.7–38.1], P = .84), or sex distribution (74.5% women and 25.5% men in both groups, P = 1.0) between groups. Furthermore, no difference in American Society of Anesthesiologists (ASA) physical status classification system score, operative laterality, or preoperative alignment was observed between groups. Patient characteristics are shown in Table 1.Figure 2Patient inclusion flow diagram.Table 1Patient demographics.Total (N = 94)AB-TKA (N = 47)MR-TKA (N = 47)P valueAge Mean (standard deviation) [range]73.5 (±7.58) [53.0, 88.7]73.5 (±7.81) [53.0, 88.5]73.5 (±7.43) [53.8, 88.7].985Sex Female70 (74.5%)35 (74.5%)35 (74.5%)1.000 Male24 (25.5%)12 (25.5%)12 (25.5%)BMI Mean (standard deviation) [range]28.7 (±4.56) [19.0, 42.0]28.8 (±4.07) [19.7, 38.1]28.6 (±5.04) [19.0, 42.0].843ASA 252 (55.3%)24 (51.1%)28 (59.6%).534 342 (44.7%)23 (48.9%)19 (40.4%)Laterality Left41 (43.6%)17 (36.2%)24 (51.1%).212 Right53 (56.4%)30 (63.8%)23 (48.9%)Preoperative alignment Neutral18 (19.1%)9 (19.1%)9 (19.1%).968 Valgus21 (22.3%)11 (23.4%)10 (21.3%) Varus55 (58.5%)27 (57.4%)28 (59.6%)ASA, American Society of Anesthesiologists. Patient inclusion flow diagram. Table 1Patient demographics.Total (N = 94)AB-TKA (N = 47)MR-TKA (N = 47)P valueAge Mean (standard deviation) [range]73.5 (±7.58) [53.0, 88.7]73.5 (±7.81) [53.0, 88.5]73.5 (±7.43) [53.8, 88.7].985Sex Female70 (74.5%)35 (74.5%)35 (74.5%)1.000 Male24 (25.5%)12 (25.5%)12 (25.5%)BMI Mean (standard deviation) [range]28.7 (±4.56) [19.0, 42.0]28.8 (±4.07) [19.7, 38.1]28.6 (±5.04) [19.0, 42.0].843ASA 252 (55.3%)24 (51.1%)28 (59.6%).534 342 (44.7%)23 (48.9%)19 (40.4%)Laterality Left41 (43.6%)17 (36.2%)24 (51.1%).212 Right53 (56.4%)30 (63.8%)23 (48.9%)Preoperative alignment Neutral18 (19.1%)9 (19.1%)9 (19.1%).968 Valgus21 (22.3%)11 (23.4%)10 (21.3%) Varus55 (58.5%)27 (57.4%)28 (59.6%)ASA, American Society of Anesthesiologists. Patient demographics. ASA, American Society of Anesthesiologists. Comparison of operative time between guidance devices The mean operative time from the first incision to incision closure was longer for the MR-TKA group compared to the AB-TKA group (109 ± 21 minutes vs 90 ± 13 minutes, respectively, P < .001; Fig. 3). The first half of MR-TKA cases completed, sorted by date, had a mean operative time of 113 ± 26 minutes, while the second half of cases had a mean operative time of 104 ± 13 minutes (P = .066).Figure 3Operative time for AB-TKA and MR-TKA. Bracket ∗∗∗ indicates P < .001. Operative time for AB-TKA and MR-TKA. Bracket ∗∗∗ indicates P < .001. Postoperative radiographic deviation from planned intraoperative alignment Three MR-TKA cases did not have follow-up lower-extremity standing radiographs available, leaving 91 total cases for analysis. Absolute average rater intraclass correlation coefficient scores of 0.98 and 0.96 were calculated for femoral resection angle and tibial resection angle, respectively, indicating “excellent” (>0.90) IRR [25]. Postoperative radiographic deviation from the intraoperative plan for femoral and tibial resection angles are shown in Figure 4. Compared to planned intraoperative alignment, postoperative deviation from the planned femoral resection angle was lower for MR-TKA compared to AB-TKA (1.20° ± 1.21° vs 1.87° ± 1.48°, respectively, P = .021). Forty of 44 (91%) MR-TKA and 39/47 (83%) AB-TKA cases were within 3 degrees of planned femoral resection (P = .26). Although postoperative deviation from planned alignment for the tibial resection angle was less in the MR-TKA group than in the AB-TKA group, the results did not reach statistical significance (1.06° ± 0.77° vs 1.39° ± 1.08, respectively, P = .093). Forty-three of 44 (98%) MR-TKA and 43/47 (91%) AB-TKA cases were within 3 degrees of planned tibial resection.Figure 4Postoperative radiographic deviation from the intraoperative plan for femoral resection angle (a) and tibial resection angle (b). Bracket ∗ indicates P < .05. NS, no significance. Postoperative radiographic deviation from the intraoperative plan for femoral resection angle (a) and tibial resection angle (b). Bracket ∗ indicates P < .05. NS, no significance. Preoperative and postoperative flexion and extension No difference in knee flexion or extension was seen between groups either preoperatively or postoperatively (Table 2). Terminal knee extension improved following both MR-TKA and AB-TKA (4.0° ± 4.1° vs 5.7° ± 4.6°, P = .062). Terminal flexion improved following both MR-TKA and AB-TKA (4.0° ± 13.0° vs 3.7° ± 12.4°, P = .927).Table 2Preoperative and postoperative range of motion (degrees).AB-TKAMR-TKAUnpaired t-testFlexion Preoperative110.2 ± 9.4111.0 ± 10.70.712 Postoperative113.6 ± 8.0114.8 ± 11.80.565 Change3.7 ± 12.44.0 ± 13.00.927 Paired t-test0.0810.051Extension Preoperative6.7 ± 5.54.7 ± 3.90.052 Postoperative1.0 ± 2.30.7 ± 1.70.526 Change5.7 ± 4.64.0 ± 4.10.062 Paired t-test<0.001<0.001 Table 2Preoperative and postoperative range of motion (degrees).AB-TKAMR-TKAUnpaired t-testFlexion Preoperative110.2 ± 9.4111.0 ± 10.70.712 Postoperative113.6 ± 8.0114.8 ± 11.80.565 Change3.7 ± 12.44.0 ± 13.00.927 Paired t-test0.0810.051Extension Preoperative6.7 ± 5.54.7 ± 3.90.052 Postoperative1.0 ± 2.30.7 ± 1.70.526 Change5.7 ± 4.64.0 ± 4.10.062 Paired t-test<0.001<0.001 Preoperative and postoperative range of motion (degrees). Postoperative complications At 3 months postoperatively, 1 of 47 (2%) MR-TKA cases and 2 of 47 (4%) AB-TKA cases had complications requiring reoperation. The one MR-TKA case required manipulation under anesthesia 2 months postoperatively to improve ROM, while both AB-TKA complications consisted of traumatic wound dehiscence requiring surgical irrigation and debridement. There was no statistically significant difference in total complications between groups (P = .56). No cases in either group required component revision. Discussion The objective of this study was to determine if TKA using mixed-reality guidance results in equivalent operative time, resection accuracy, and short-term clinical outcomes compared to accelerometer-based guidance. We found that MR-TKA is associated with a longer operative time (average 19 minutes) and higher femoral resection accuracy (0.67 degrees less deviation from plan) compared to AB-TKA. No statistically significant differences in ROM or complications were observed at 3 months, suggesting a comparable short-term safety profile. Together, these findings suggest that MR-TKA offers comparable results to AB-TKA, with nuances in operative time and resection accuracy that individual practices may consider when evaluating TKA navigation options. Operative time may be elevated in this study due to the novel nature of this recently Food and Drug Administration-cleared device and an associated learning curve, discussed further below. Device learning curve and registration requirements The increased operative time for MR-TKA observed in this study is likely attributed to both a learning curve and increased device registration requirements. Prior studies investigating the learning curve for adoption of AB-TKA found that it takes approximately 10-20 cases before a novice has a similar operating time to an experienced surgeon [15,16]. We found the mean operative time was 9 minutes faster for the second half of MR-TKA cases performed in this study compared to the first, indicating possible learning curve effects. However, the second half of MR-TKA cases were still on average 14 minutes longer than AB-TKA, which may be due to the additional steps required to register optical tracking arrays [26] and potentially longer learning curve for MR-TKA. In AB-TKA, femoral registration is based solely on the distal femoral condyles and the hip center of rotation. To perform tibial resection using AB-TKA, the lateral and medial malleoli are registered to approximate the mechanical axis of the tibia, allowing the surgeon to adjust the tibial jig relative to the position of these 2 tracked points [27]. Registration of MR-TKA guidance is relatively more complex, as there are 14 distinct landmarks to be captured during registration, 7 each for the femur and tibia [28]. In a randomized controlled study of 50 patients undergoing TKA, Swamy et al. found that handheld, accelerometer-based guidance, the AB-TKA guidance system used in this study, was significantly faster (average 55 minutes) than traditional computer-assisted guidance (average 62 minutes) that required increased registration [29]. We approximate that both the mixed-reality guidance device and accelerometer-based guidance device each required approximately 3-5 minutes to set up, including calibration. This was primarily performed by the surgical technologist and did not prolong operative time. The mixed-reality headset required additional optical registration, which was performed by the company vendor prior to the procedure. Clinical relevance of increased resection accuracy One intended benefit of TKA guidance systems utilizing imageless optical navigation is that resection accuracy may be improved, as historically seen in both robotic- [30] and computer-assisted studies [31]. In this study, we found that MR-TKA is on average 0.67 degrees more accurate in femoral resection accuracy (1.20° mean deviation from plan) compared to AB-TKA (1.87°). Resection deviation of 3 degrees has been established in the literature as a threshold for increased risk of hardware loosening complications [32]. For the average case, an improved accuracy of 0.67 degrees is not likely to influence patient outcomes or device longevity. However, increased accuracy may reduce the number of outlier procedures resulting in greater than 3 degrees, which is reflected in our findings (91% of MR-TKA cases were found within 3 degrees of planned femoral resection, while only 83% of AB-TKA cases were found within this threshold) and decreases the chances of propagating errors. Similarly, although the numerical difference in tibial resection accuracy did not meet statistical significance, tibial resection using MR-TKA also trended toward more accurate than that of AB-TKA, reflected by fewer deviations outside the 3 degree threshold (98% MR-TKA and 91% AB-TKA). In future studies, postoperative femoral rotation may be assessed to further elucidate differences in accuracy between the devices and help determine whether increased accuracy is provided by tracking of femoral rotation by the MR-TKA guidance system. Practical lessons during early adoption of mixed-reality TKA In the adoption of this novel MR-TKA guidance system, we learned several practical lessons outside of our study aims. In a teaching hospital with trainees such as residents or fellows, MR-TKA can be live-streamed to provide a view of the surgery from the point of view of the surgeon, which may help improve understanding of the intraoperative decision-making process. Conversely, a downside of current MR-TKA platforms, however, is that a mixed-reality headset cannot be easily transferred from one surgeon to another, particularly in a sterile setting using surgical helmet systems, making it difficult for other surgical team members to directly participate. Finally, the optical arrays themselves can be a liability in a surgery, as loosening of the array may invalidate the tracking and force the surgeon to recalibrate or switch to a different guidance system or manual instrumentation. Study limitations This study has several limitations inherent to retrospective designs that should be considered when interpreting the data, including limited follow-up, incomplete or missing data, selection and observer bias, and a relatively small sample size. Clinical follow-up for this study was limited to 3 months, which precludes assessment of long-term factors such as patient satisfaction or implant survivorship. Additionally, postoperative standing radiographs were taken in a 2-dimensional anterior-posterior view, which is prone to measurement inaccuracies if the knee joint is not properly aligned with the coronal plane [7]. In future studies, the preoperative patient population may be examined more precisely by assessing both joint line obliquity and hip-knee-ankle axis to evaluate coronal plane alignment of the knee. Future studies with larger cohorts may also benefit from expanded tracking of nonoperative complications, such as deep vein thrombosis, pulmonary embolism, and other complications not requiring reoperation. Furthermore, the single-institution setting and two-surgeon cohort may limit external validity. The historical control with accelerometer-based guidance was available from only one surgeon, as the other surgeon does not use that device. Future evaluations with patient-reported outcomes will be critical to evaluate if there is a difference in functional outcomes or patient satisfaction associated with the different guidance systems. Conclusions MR-TKA was associated with modestly higher femoral resection accuracy, but longer operative time, compared to AB-TKA, possibly due to additional calibration steps and a longer learning curve. There was no difference in clinically observed knee ROM or complications at 3-month follow-up. Long-term studies are needed to compare patient-reported outcomes and implant survivorship between the 2 techniques. CRediT authorship contribution statement Michael A. Hewitt: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Software, Resources, Project administration, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Catelyn A. Woelfle: Writing – review & editing, Writing – original draft, Validation, Methodology, Data curation. Anthony H. Zou: Writing – review & editing, Writing – original draft, Validation, Methodology, Data curation. Wyatt H. Buchalter: Writing – review & editing, Writing – original draft, Validation, Methodology, Formal analysis. Alexander L. Neuwirth: Writing – review & editing, Writing – original draft, Supervision, Methodology, Investigation, Formal analysis. Carl L. Herndon: Writing – review & editing, Writing – original draft, Supervision, Methodology, Investigation, Formal analysis. H. John Cooper: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Resources, Project administration, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Nana O. Sarpong: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Resources, Project administration, Methodology, Investigation, Data curation, Conceptualization. Conflict of interest A.L. Neuwirth has stock or stock options in Sylke and received other financial or material support from Smith and Nephew Orthopaedics. C.L. Herndon is a paid consultant for Johnson & Johnson Med Tech, Medtronic, and PolarisAR and is a board member/committee appointments for Journal of Arthroplasty. C.A. Woelfle is a paid consultant for PolarisAR and received other financial or material support from PolarisAR. H.J. Cooper is a paid consultant for PolarisAR; has stock or stock options in PolarisAR; received other financial or material support from Solventum and Zimmer Biomet Holdings, Inc; and is a board member/committee appointments for AAHKS Research Committee. N.O. Sarpong is a paid consultant for PolarisAR; has stock or stock options in PolarisAR; received other financial or material support from Zimmer Biomet Holdings, Inc. W.H. Buchalter is a paid consultant for PolarisAR; all other authors declare no potential conflicts of interest. For full disclosure statements refer to https://doi.org/10.1016/j.artd.2026.102034.