Is it feasible to develop a supervised learning algorithm incorporating spinopelvic mobility to predict impingement in patients undergoing total hip arthroplasty?

Aims: Precise implant positioning, tailored to individual spinopelvic biomechanics and phenotype, is paramount for stability in total hip arthroplasty (THA). Despite a few studies on instability prediction, there is a notable gap in research utilizing artificial intelligence (AI). The objective of our pilot study was to evaluate the feasibility of developing an AI algorithm tailored to individual spinopelvic mechanics and patient phenotype for predicting impingement. Methods: This international, multicentre prospective cohort study across two centres encompassed 157 adults undergoing primary robotic arm-assisted THA. Impingement during specific flexion and extension stances was identified using the virtual range of motion (ROM) tool of the robotic software. The primary AI model, the Light Gradient-Boosting Machine (LGBM), used tabular data to predict impingement presence, direction (flexion or extension), and type. A secondary model integrating tabular data with plain anteroposterior pelvis radiographs was evaluated to assess for any potential enhancement in prediction accuracy. Results: We identified nine predictors from an analysis of baseline spinopelvic characteristics and surgical planning parameters. Using fivefold cross-validation, the LGBM achieved 70.2% impingement prediction accuracy. With impingement data, the LGBM estimated direction with 85% accuracy, while the support vector machine (SVM) determined impingement type with 72.9% accuracy. After integrating imaging data with a multilayer perceptron (tabular) and a convolutional neural network (radiograph), the LGBM's prediction was 68.1%. Both combined and LGBM-only had similar impingement direction prediction rates (around 84.5%). Conclusion: This study is a pioneering effort in leveraging AI for impingement prediction in THA, utilizing a comprehensive, real-world clinical dataset. Our machine-learning algorithm demonstrated promising accuracy in predicting impingement, its type, and direction. While the addition of imaging data to our deep-learning algorithm did not boost accuracy, the potential for refined annotations, such as landmark markings, offers avenues for future enhancement. Prior to clinical integration, external validation and larger-scale testing of this algorithm are essential.

Strong Correlation Between Standing Long-Leg Radiographs and CT Scans in Measuring Coronal Knee Alignment.

BACKGROUND: The objective of this study was to evaluate the correlation in measurements of the lower-limb coronal alignment between long-leg radiographs (LLRs) and computed tomography (CT) scanograms that were made during preoperative planning for robotic-arm-assisted knee arthroplasty. On the basis of published evidence demonstrating a good correlation between these imaging modalities in measuring the lower-limb mechanical axis, we hypothesized that there would be no significant differences between the 2 in the present study. METHODS: This multicenter cohort study across 3 tertiary centers included 300 patients undergoing primary robotic-arm-assisted total knee arthroplasty (TKA) or unicompartmental knee arthroplasty (UKA) for whom LLRs and CT scanograms were available preoperatively. The study involved measuring the medial proximal tibial angle (MPTA), lateral distal femoral angle (LDFA), hip-knee-ankle angle (HKA), joint line obliquity (JLO), joint-line convergence angle (JLCA), and arithmetic HKA (aHKA). The aHKA represents a method for estimating constitutional alignment using angles that are unaffected by joint-space narrowing. RESULTS: Strong correlations (p < 0.001) between the imaging modalities were found for the HKA (correlation coefficient, 0.912), aHKA (0.883), MPTA (0.820), LDFA (0.871), and JLO (0.778). A weaker correlation was observed for the JLCA in valgus knees as compared with varus knees (Spearman coefficients, 0.412 and 0.518, respectively). Regression models demonstrated that the degree of agreement was associated with the preoperative intra-articular deformity and the positioning of the lower limb during the CT scan (i.e., the lower-limb rotational angle). An initial JLCA within ±5° was associated with higher agreement. CONCLUSIONS: We observed a strong correlation between LLRs and CT scanograms that were made during the preoperative planning stage of robotic-arm-assisted knee arthroplasty, implying that CT scanograms can reliably be utilized to estimate the coronal alignment of the knee, potentially replacing the need for LLRs. Nevertheless, to attain a higher degree of agreement, it is crucial to ensure appropriate radiographic positioning of the lower limb. Additionally, surgeons must remain vigilant regarding potential discrepancies in cases involving substantial deformities. LEVEL OF EVIDENCE: Prognostic Level II. See Instructions for Authors for a complete description of levels of evidence.

Functional Component Positioning in Total Hip Arthroplasty and the Role of Robotic-Arm Assistance in Addressing Spinopelvic Pathology.

Hip, spine, and pelvis function as a unified kinetic chain. Any spinal pathology, results in compensatory changes in the other components to accommodate for the reduced spinopelvic motion. The complex relationship between spinopelvic mobility and component positioning in total hip arthroplasty presents a challenge in achieving functional implant positioning. Patients with spinal pathology, especially those with stiff spines and little change in sacral slope, are at high instability risk. In this challenging subgroup, robotic-arm assistance enables the execution of a patient specific plan, avoiding impingement and maximizing range of motion; especially utilizing virtual range of motion to dynamically assess impingement.

Functional implant positioning in total hip arthroplasty and the role of robotic-arm assistance.

INTRODUCTION: Accurate implant positioning, tailored to the phenotype and unique biomechanics of each patient is the single most important objective in achieving stability in THA and maximise range of motion. The spine-pelvis-hip construct functions as a single unit adapting to postural changes. It is widely accepted in the literature that no universaltarget exists and variations in spinopelvic mobility mandate adjustments to the surgical plan; thus bringing to the fore the concept of personalised, functional component positioning. METHODS: This manuscript aims to outline the challenges posed by spinopelvic imbalance and present a reproducible, stepwise approach to achieve functional-component positioning. We also present the one-year functional outcomes and Patient Reported Outcome Measures of a prospective cohort operated with this technique. RESULTS AND CONCLUSION: Robotic-arm assisted Total Hip Arthroplasty has facilitated enhanced planning based on the patient's phenotype and evidence suggests it results in more reproducible and accurate implant positioning. Preservation of offset, avoiding leg-length discrepancy, accurate restoration of the centre of rotation and accomplishing the combinedversion target are very important parameters in Total Hip Arthroplasty that affect post-operative implant longevity, patient satisfaction and clinical outcomes.

Robotic-Arm-Assisted Total Hip Arthroplasty: A Review of the Workflow, Outcomes and Its Role in Addressing the Challenge of Spinopelvic Imbalance.

Robotic-arm-assisted total hip arthroplasty (RoTHA) offers the opportunity to improve the implant positioning and restoration of native hip mechanics. The concept of individualised, functional implant positioning and how it relates to spinopelvic imbalance is an important yet rather novel consideration in THA. There is mounting evidence that a significant percentage of dislocations occur within the perceived "safe zones"; hence, in the challenging subset of patients with a stiff spinopelvic construct, it is imperative to employ individualised component positioning based on the patients' phenotype. Restoring the native centre of rotation, preserving offset, achieving the desired combined anteversion and avoiding leg length inequality are all very important surgeon-controlled variables that have been shown to be associated with postoperative outcomes. The latest version of the software has a feature of virtual range of motion (VROM), which preoperatively identifies potential dynamic causes of impingement that can cause instability. This review presents the workflow of RoTHA, especially focusing on pragmatic solutions to tackle the challenge of spinopelvic imbalance. Furthermore, it presents an overview of the existing evidence concerning RoTHA and touches upon future direction.