Classifying Toe Walking Gait Patterns Among Children Diagnosed With Idiopathic Toe Walking Using Wearable Sensors and Machine Learning Algorithms

Idiopathic toe walking (ITW) is a gait abnormality in which children touch at initial contact and demonstrate limited or no heel throughout the cycle. Toe results poor balance, increased risk of falling, developmental delays among children. Identifying steps during can facilitate targeted intervention diagnosed with ITW. With recent advances wearable sensing, communication technologies, machine learning, new avenues managing behavior are feasible. In this study, we investigate capabilities Machine Learning (ML) algorithms identifying foot (heel strike versus strike) utilizing body sensors. Thirty-six (Age 9.4±2.8 years) ITW participated study. Six ML algorithms, consisting Support Vector Machines (SVM), decision tree (DT), random forest (RF), K-nearest neighbors (KNN), Multi-layer Perceptron (MLP), Gaussian process (GP), could successfully classify patterns We found that simple KNN algorithm resulted highest accuracy 92.92% F1-score 93.20% to differentiate best when using all four also higher variability sacral vertical accelerations Accurate quantification clinical applications critical for assessing rehabilitation progress designing interventions