Diagnosing Huntington's Disease Through Gait Dynamics

This study proposes an automatic method for identifying Huntington's disease using features extracted from gait signals derived from force-sensitive resistors. Features were extracted using metrics of fluctuation magnitude and fluctuation dynamics, obtained from a detrended Fluctuation Analysis (DFA). In the classification, five machine learning algorithms (Support Vector Machines (SVM), K-Nearest Neighbor (KNN), Naive Bayes (NB), Linear Discriminant Analysis (LDA) and Decision Tree (DT)) were compared by the leave-one-out cross-validation method. Our experiments showed that SVM and DT provided the best results, achieving an average accuracy of 100.0%, representing an improvement compared to other results in the literature, and proving the effectiveness of the proposed method.