Predicting robotic-assisted total knee arthroplasty operating time

Artificial Intelligence in Commercial Industry: Serving the End-to-End Patient Experience Across the Digital Ecosystem

The purpose of this article is to evaluate the application of AI from the perspective of orthopaedic industry with respect to the specific opportunities offered by AI. It is clear that AI has the potential to impact the entire continuum of musculoskeletal and orthopaedic care. The following areas may experience improvements from integrating AI into surgical applications: surgical trainees can learn more easily at lower costs in extended reality simulations; physicians can receive support in decision making and case planning; efficiencies can be driven with improved case management and hospital episodes; performing surgery - which until recently was the only element industry engaged with - can benefit from intra-operative AI derived inputs; finally, post-operative care can be tailored to the individual patient and their circumstances. AI delivers the potential for industry to offer valuable augments to patient experience and enhanced surgical insights along the digital episode of care. However, the true value is in considering not just how AI can be applied in each silo but across the patient's entire continuum of care. This opportunity was first opened with the advent of robotics. The data derived from the robotic systems has added something akin to a black box flight recorder to the operation, which now offers two critical outcomes for industry. First, together we can now start to stitch pre-operative elements like demographics, morphological phenotyping, and pathology that can be integrated with intra-operative elements to produce surgical plans and on-the-fly anatomic data like ligament tension. Second, post-operative elements such as recovery protocols and outcomes can be considered through the lens of the intra-operative experience. In forming this bridge, AI can accelerate the development of a truly integrated digital ecosystem, facilitating a shift from providing implants to providing patient experience pathways.

Surgical factors play a critical role in predicting functional outcomes using machine learning in robotic-assisted total knee arthroplasty

PURPOSE

Predictive models help determine predictive factors necessary to improve functional outcomes after total knee arthroplasty (TKA). However, no study has assessed predictive models for functional outcomes after TKA based on the new concepts of personalised surgery and new technologies. This study aimed to develop and evaluate predictive modelling approaches to predict the achievement of minimal clinically important difference (MCID) in patient-reported outcome measures (PROMs) 1 year after TKA.

METHODS

Four hundred thirty robotic-assisted TKAs were analysed in this retrospective study. The mean age was 67.9 ± 7.9 years; the mean body mass index (BMI) was 32.0 ± 6.8 kg/m2. The following PROMs were collected preoperatively and 1-year postoperatively: knee injury and osteoarthritis outcome score for joint replacement, Western Ontario and McMaster Universities osteoarthritis index (WOMAC) Function, WOMAC Pain. Demographic data, preoperative CT scan, implant size, implant position on the robotic system and characteristics of the joint replacement procedure were selected as predictive variables. Four machine learning algorithms were trained to predict the MCID status at 1-year post-TKA for each PROM survey. 'No MCID' was chosen as the target. Models were evaluated by class discrimination (F1-score) and area under the receiver operating characteristic curve (ROC-AUC).

RESULTS

The best-performing model was ridge logistic regression for WOMAC Function (area under the curve [AUC]  = 0.80, F1 = 0.48, sensitivity = 0.79, specificity = 0.62). Variables most strongly contributing to not achieving MCID status were preoperative PROMs, high BMI and femoral resection depth (posterior and distal), supporting functional positioning principles. Conversely, variables contributing to a positive outcome (achieving MCID) were medial/lateral alignment of the tibial component, whether the procedure was an outpatient surgery and whether the patient received managed Medicare insurance.

CONCLUSION

The most predictive variables included preoperative PROMs, BMI and surgical planning. The surgical predictive variables were valgus femoral alignment and femoral rotation, reflecting the benefits of personalised surgery. Including surgical variables in predictive models for functional outcomes after TKA should guide clinical and surgical decision-making for every patient.

LEVEL OF EVIDENCE

Level III.

ISTA Award 2023: Toward functional reconstruction of the pre-diseased state in total knee arthroplasty

Aims

The surgical target for optimal implant positioning in robotic-assisted total knee arthroplasty remains the subject of ongoing discussion. One of the proposed targets is to recreate the knee's functional behaviour as per its pre-diseased state. The aim of this study was to optimize implant positioning, starting from mechanical alignment (MA), toward restoring the pre-diseased status, including ligament strain and kinematic patterns, in a patient population.

Methods

We used an active appearance model-based approach to segment the preoperative CT of 21 osteoarthritic patients, which identified the osteophyte-free surfaces and estimated cartilage from the segmented bones; these geometries were used to construct patient-specific musculoskeletal models of the pre-diseased knee. Subsequently, implantations were simulated using the MA method, and a previously developed optimization technique was employed to find the optimal implant position that minimized the root mean square deviation between pre-diseased and postoperative ligament strains and kinematics.

Results

There were evident biomechanical differences between the simulated patient models, but also trends that appeared reproducible at the population level. Optimizing the implant position significantly reduced the maximum observed strain root mean square deviations within the cohort from 36.5% to below 5.3% for all but the anterolateral ligament; and concomitantly reduced the kinematic deviations from 3.8 mm (SD 1.7) and 4.7° (SD 1.9°) with MA to 2.7 mm (SD 1.4) and 3.7° (SD 1.9°) relative to the pre-diseased state. To achieve this, the femoral component consistently required translational adjustments in the anterior, lateral, and proximal directions, while the tibial component required a more posterior slope and varus rotation in most cases.

Conclusion

These findings confirm that MA-induced biomechanical alterations relative to the pre-diseased state can be reduced by optimizing the implant position, and may have implications to further advance pre-planning in robotic-assisted surgery in order to restore pre-diseased knee function.

Utilization of Machine Learning Models to More Accurately Predict Case Duration in Primary Total Joint Arthroplasty

BACKGROUND

Accurate operative scheduling is essential for the appropriation of operating room esources. We sought to implement a machine learning model to predict primary total hip arthroplasty (THA) and total knee arthroplasty (TKA) case time.

METHODS

A total of 10,590 THAs and 12,179 TKAs between July 2017 and December 2022 were retrospectively identified. Cases were chronologically divided into training, validation, and test sets. The test set cohort included 1,588 TKAs and 1,204 THAs. There were four ML algorithms developed: linear ridge regression (LR), random forest, XGBoost, and explainable boosting machine. Each model's case time estimate was compared to the scheduled estimate measured in 15-minute "wait" time blocks ("underbooking") and "excess" time blocks ("overbooking"). Surgical case time was recorded, and SHAP values were assigned to patient characteristics, surgical information, and the patient's medical condition to understand feature importance.

RESULTS

The most predictive model input was "median previous 30 procedure case times." The XGBoost model outperformed the other models in predicting both TKA and THA case times. The model reduced TKA 'excess time blocks' by 85 blocks (P < 0.001) and 'wait time blocks' by 96 blocks (P < 0.001). The model did not significantly reduce 'excess time blocks' in THA (P = 0.89) but did significantly reduce 'wait time blocks' by 134 blocks (P < 0.001). In total, the model improved TKA operative booking by 181 blocks (2,715 minutes) and THA operative booking by 138 blocks (2,070 minutes).

CONCLUSIONS

Machine learning outperformed a traditional method of scheduling total joint arthroplasty cases. The median time of the prior 30 surgical cases was the most influential on scheduling case time accuracy. As ML models improve, surgeons should consider ML utilization in case scheduling; however, prior 30 surgical cases may serve as an adequate alternative.

Clinical validation of a deep learning-based approach for preoperative decision-making in implant size for total knee arthroplasty

BACKGROUND

Orthopedic surgeons use manual measurements, acetate templating, and dedicated software to determine the appropriate implant size for total knee arthroplasty (TKA). This study aimed to use deep learning (DL) to assist in deciding the femoral and tibial implant sizes without manual manipulation and to evaluate the clinical validity of the DL decision by comparing it with conventional manual procedures.

METHODS

Two types of DL were used to detect the femoral and tibial regions using the You Only Look Once algorithm model and to determine the implant size from the detected regions using convolutional neural network. An experienced surgeon predicted the implant size for 234 patient cases using manual procedures, and the DL model also predicted the implant sizes for the same cases.

RESULTS

The exact accuracies of the surgeon's template were 61.54% and 68.38% for predicting femoral and tibial implant sizes, respectively. Meanwhile, the proposed DL model reported exact accuracies of 89.32% and 90.60% for femoral and tibial implant sizes, respectively. The accuracy ± 1 levels of the surgeon and proposed DL model were 97.44% and 97.86%, respectively, for the femoral implant size and 98.72% for both the surgeon and proposed DL model for the tibial implant size.

CONCLUSION

The observed differences and higher agreement levels achieved by the proposed DL model demonstrate its potential as a valuable tool in preoperative decision-making for TKA. By providing accurate predictions of implant size, the proposed DL model has the potential to optimize implant selection, leading to improved surgical outcomes.

Lack of Validity of Absolute Percentage Errors in Estimated Operating Room Case Durations as a Measure of Operating Room Performance: A Focused Narrative Review

Commonly reported end points for operating room (OR) and surgical scheduling performance are the percentages of estimated OR times whose absolute values differ from the actual OR times by ≥15%, or by various intervals from ≥5 to ≥60 minutes. We show that these metrics are invalid assessments of OR performance. Specifically, from 19 relevant articles, multiple OR management decisions that would increase OR efficiency or productivity would also increase the absolute percentage error of the estimated case durations. Instead, OR managers should check the mean bias of estimated OR times (ie, systematic underestimation or overestimation), a valid and reliable metric.

Factors Affecting Operating Room Scheduling Accuracy for Primary and Revision Total Knee Arthroplasty: A Retrospective Study

BACKGROUND

Optimizing operating room (OR) scheduling accuracy is important for improving OR efficiency and maximizing value of total knee arthroplasty (TKA). However, data on factors that may impact TKA OR scheduling accuracy are limited.

MATERIALS AND METHODS

A retrospective review of 7655 knee arthroplasties (6999 primary TKAs and 656 revision TKAs) performed between January 2020 and May 2023 was conducted. Patient baseline characteristics, surgeon experience (years in practice), as well as actual vs scheduled OR times were collected. Actual OR times that were at least 15% shorter or longer than scheduled OR times were considered to be clinically important. Logistic regression analyses were employed to assess the influence of specific patient and surgeon factors on OR scheduling inaccuracies.

RESULTS

Using adjusted odds ratio, patients with primary TKA who had a lower body mass index (P<.001) were independently associated with overestimation of scheduled surgical time. Conversely, younger age (P<.001), afternoon procedure start time (P<.001), surgeons with less than 10 years of experience (P=.037), and higher patient body mass index (P<.001) were associated with underestimation of scheduled surgical time. For revision TKA, female sex (P=.021) and morning procedure start time (P=.038) were associated with overestimation of scheduled surgical time, while surgeons with less than 10 years of experience (P=.014) and patients who underwent spinal/epidural/block anesthesia (P=.038) were associated with underestimation of scheduled surgical time.

CONCLUSION

This study highlights patient, surgeon, and intraoperative variables that impact the accuracy of scheduling for TKA procedures. Health systems should take these variables into consideration when creating OR schedules to fully optimize resources and available space. [Orthopedics. 2024;47(5):313-319.].

Initial Learning Curve for Robot-Assisted Total Knee Arthroplasty in a Dedicated Orthopedics Center

Background and objectives: Our study aimed to assess the learning curve for robot-assisted (RA) total knee arthroplasty (TKA) in our hospital, compare operative times between RA-TKAs and manual TKAs, and assess the early complications rate between the two approaches. Methods: We included 39 patients who underwent RA-TKA and 45 control patients subjected to manual TKA in the same period and operated on by the same surgical staff. We collected demographic and patient-related data to assess potential differences between the two groups. Results: No statistical differences were recorded in regard to age, BMI, sex, Kellgren-Lawrence classification, or limb alignment between patients undergoing RA-TKA and manual TKA, respectively. Three surgeons transitioned from the learning to the proficiency phase in our study after a number of 6, 4, and 3 cases, respectively. The overall operative time for the learning phase was 111.54 ± 20.45 min, significantly longer compared to the average of 86.43 ± 19.09 min in the proficiency phase (p = 0.0154) and 80.56 ± 17.03 min for manual TKAs (p < 0.0001). No statistically significant difference was recorded between the global operative time for the proficiency phase TKAs versus the controls. No major complications were recorded in either RA-TKA or manual TKA groups. Conclusions: Our results suggest that experienced surgeons may adopt RA-TKA using this platform and quickly adapt without significant complications.

Prediction of early functional outcomes in patients after robotic-assisted total knee arthroplasty: a nomogram prediction model

BACKGROUND

Robotic-assisted total knee arthroplasty (RA-TKA) is becoming more and more popular as a treatment option for advanced knee diseases due to its potential to reduce operator-induced errors. However, the development of accurate prediction models for postoperative outcomes is challenging. This study aimed to develop a nomogram model to predict the likelihood of achieving a beneficial functional outcome. The beneficial outcome is defined as a postoperative improvement of the functional Knee Society Score (fKSS) of more than 10 points, 3 months after RA-TKA by early collection and analysis of possible predictors.

METHODS

This is a retrospective study on 171 patients who underwent unilateral RA-TKA at our hospital. The collected data included demographic information, preoperative imaging data, surgical data, and preoperative and postoperative scale scores. Participants were randomly divided into a training set ( N =120) and a test set ( N =51). Univariate and multivariate logistic regression analyses were employed to screen for relevant factors. Variance inflation factor was used to investigate for variable collinearity. The accuracy and stability of the models were evaluated using calibration curves with the Hosmer-Lemeshow goodness-of-fit test, consistency index and receiver operating characteristic curves.

RESULTS

Predictors of the nomogram included preoperative hip-knee-ankle angle deviation, preoperative 10-cm Visual Analogue Scale score, preoperative fKSS score and preoperative range of motion. Collinearity analysis with demonstrated no collinearity among the variables. The consistency index values for the training and test sets were 0.908 and 0.902, respectively. Finally, the area under the receiver operating characteristic curve was 0.908 (95% CI 0.846-0.971) in the training set and 0.902 (95% CI 0.806-0.998) in the test set.

CONCLUSION

A nomogram model was designed hereby aiming to predict the functional outcome 3 months after RA-TKA in patients. Rigorous validation showed that the model is robust and reliable. The identified key predictors include preoperative hip-knee-ankle angle deviation, preoperative visual analogue scale score, preoperative fKSS score, and preoperative range of motion. These findings have major implications for improving therapeutic interventions and informing clinical decision-making in patients undergoing RA-TKA.

Pre-Planning the Surgical Target for Optimal Implant Positioning in Robotic-Assisted Total Knee Arthroplasty

Robotic-assisted total knee arthroplasty can attain highly accurate implantation. However, the target for optimal positioning of the components remains debatable. One of the proposed targets is to recreate the functional status of the pre-diseased knee. The aim of this study was to demonstrate the feasibility of reproducing the pre-diseased kinematics and strains of the ligaments and, subsequently, use that information to optimize the position of the femoral and tibial components. For this purpose, we segmented the pre-operative computed tomography of one patient with knee osteoarthritis using an image-based statistical shape model and built a patient-specific musculoskeletal model of the pre-diseased knee. This model was initially implanted with a cruciate-retaining total knee system according to mechanical alignment principles; and an optimization algorithm was then configured seeking the optimal position of the components that minimized the root-mean-square deviation between the pre-diseased and post-operative kinematics and/or ligament strains. With concurrent optimization for kinematics and ligament strains, we managed to reduce the deviations from 2.4 ± 1.4 mm (translations) and 2.7 ± 0.7° (rotations) with mechanical alignment to 1.1 ± 0.5 mm and 1.1 ± 0.6°, and the strains from 6.5% to lower than 3.2% over all the ligaments. These findings confirm that adjusting the implant position from the initial plan allows for a closer match with the pre-diseased biomechanical situation, which can be utilized to optimize the pre-planning of robotic-assisted surgery.

Improving Resource Utilization for Arthroplasty Care by Leveraging Machine Learning and Optimization: A Systematic Review

Background

There is a growing demand for total joint arthroplasty (TJA) surgery. The applications of machine learning (ML), mathematical optimization, and computer simulation have the potential to improve efficiency of TJA care delivery through outcome prediction and surgical scheduling optimization, easing the burden on health-care systems. The purpose of this study was to evaluate strategies using advances in analytics and computational modeling that may improve planning and the overall efficiency of TJA care.

Methods

A systematic review including MEDLINE, Embase, and IEEE Xplore databases was completed from inception to October 3, 2022, for identification of studies generating ML models for TJA length of stay, duration of surgery, and hospital readmission prediction. A scoping review of optimization strategies in elective surgical scheduling was also conducted.

Results

Twenty studies were included for evaluating ML predictions and 17 in the scoping review of scheduling optimization. Among studies generating linear or logistic control models alongside ML models, only 1 found a control model to outperform its ML counterpart. Furthermore, neural networks performed superior to or at the same level as conventional ML models in all but 1 study. Implementation of mathematical and simulation strategies improved the optimization efficiency when compared to traditional scheduling methods at the operational level.

Conclusions

High-performing predictive ML-based models have been developed for TJA, as have mathematical strategies for elective surgical scheduling optimization. By leveraging artificial intelligence for outcome prediction and surgical optimization, there exist greater opportunities for improved resource utilization and cost-savings in TJA than when using traditional modeling and scheduling methods.