Evaluation of machine learning methods to predict knee loading from the movement of body segments

Artificial Intelligence for Clinical Gait Diagnostics of Knee Osteoarthritis: An Evidence - based Review and Analysis

“…Furthermore, the incidence of symptomatic knee OA is increasing due to the ageing population and obesity, particularly in developed countries [3]. Quantitative analysis of the gait can help clinicians diagnose knee-related conditions by recognising deviations from physiological gait, e.g., knee adduction moment (KAM) in patients with knee OA [5,6]. However, analysis of mocap data heavily relies on adequate experimental setup, pre-and post-processing methods, e.g., gap filling between marker trajectories, adequate band-pass filter design and thresholds.…”

“…Whilst unsupervised AI learning-based algorithms, such as the self-organising map (SOM) [11,12], Random Forest (RF) [13,14], can classify gait data used as inputs without preliminarily knowing their true classes, supervised classifiers instead, such as the multi-layer perceptron (MLP) [15], the radial basis function (RBF) networks [16] and the Support Vector Machine (SVM) [14] require that the true classes of the input data are preliminarily known. Amongst AI-based methods, Machine Learning (ML) [7,[17][18][19] and Artificial Neural Networks (ANN) have proven to be accurate when dealing with gait-related data on patients with knee OA [2,5,6,20,21].…”

“…Instead, in this study not only a qualitative analysis of previous research findings was carried, but also a quantitative one was performed using a novel evidence-based analysis approach in AI for clinical gait diagnostics, which yielded two eligible studies that compared the classification performance of ML-and ANN-based methods directly, i.e., in the same study. Aljaaf et al [6] were able to capture underlying gait patterns from the movement of body segments and quantified the knee adduction moment of the ankle, knee (KAM), hip and pelvis from their corresponding Euler angles during a single gait cycle. As compared to other ANN-based techniques, MLP was the most accurate algorithm tested (r 2 =0.86, root mean squared error (RMSE)=0.07).…”

“…In selected papers that directly compared classification outcomes between machine learning (ML)-and artificial neural networks (ANN)based architectures (N=2) [6,18], ML-based algorithms were found to consistently have a higher accuracy and a lower standard deviation than those of ANN, thus being more stable in dealing with gait data ( Table 3). The reduced performance in ANN-based algorithms may be partly explained by the re-sampling occurring within ANN-based architectures, such as the multi-layer perceptron (MLP), where weights and biases are adjusted iteratively, and so the data is continuously resampled until the mean squared error drops below a preset threshold that is deemed acceptable to stop training the ANN.…”

“…Therefore, the lack of publication bias supports the reliability of the above-mentioned conclusions derived from analysing the results reported in Table 3. Moreover, Table 3 seems to suggest that ANN can be used when dealing with gait data collected on patients with knee OA [18], whilst ML seems to generalise to patients with any other knee-related conditions better [6,18].…”