Unraveling Freezing of Gait in Parkinson’s Disease: Datasets, Algorithms, and Open-Source Solutions

Freezing of gait (FoG) is a complex and debilitating symptom of Parkinson’s disease (PD), consisting of sudden motor block. FoG leads to an increased risk of falls and related injuries, as well as a reduction in autonomy and quality of life. Over the past two decades, a wide range of technological solutions has been proposed for the automatic detection of FoG, with heterogeneous experimental procedures, sensing technologies, algorithms, and results. Motion data can be collected using wearable sensors, environmental sensors, cameras, and radio sensors. Data can be analyzed using varying complex approaches, from simple threshold-based methods to more advanced machine and deep learning models. The encouraging results suggest that accurate and timely FoG detection can be achieved in various contexts. This chapter explores the broad and heterogeneous set of sensing systems and algorithms developed for effective FoG detection. Furthermore, public datasets and algorithms are summarized and comprehensively discussed. Finally, current limitations are identified, and possible solutions are provided. Overall, despite the research community’s constant efforts, larger and more comprehensive datasets are needed to identify the best digital solution for the continuous detection and characterization of FoG. In addition, semi-supervised or unsupervised approaches are needed to handle unconstrained monitoring in daily life. Finally, sharing data and algorithms with the research community is highly recommended. Future studies can use the information provided in this article to push research toward solutions aimed at efficient and effective FoG recognition methods that are ready to work continuously in real life, thus improving the quality of life of people with PD.