Segmentation of Gait Sequences using Inertial Sensor Data in Hereditary Spastic Paraplegia

Gait analysis is an important tool for diagnosis, monitoring and treatment of neurological diseases. Among these are hereditary spastic paraplegias (HSPs) whose main characteristic is heterogeneous gait disturbance. So far HSP gait has been analysed in a limited number of studies, and within a laboratory set up only. Although the rarity of orphan diseases often limits larger scale studies, the investigation of these diseases is still important, not only to the affect population, but also for other diseases which share gait characteristics. In this paper we use foot-mounted inertial measurement units (IMU) as a mobile solution to measure the gait of 21 HSP patients while performing a 4 by 10 m walk at self-selected pace. Two algorithms common to other gait analysis solutions, the hidden Markov model (HMM) and dynamic time warping (DTW), were applied to these signals in order to investigate their effectiveness when faced with the heterogeneous nature and range of foot strike techniques of HSP gait, sometimes even lacking a heel strike. Using a nested cross validation for parameter choice and validation, the HMM was found to be superior for segmentation purposes with a mean segmentation error of 0.10 +/- 0.05 s. Stride segmentation of such a diverse dataset is the first step towards creating a clinically relevant system which could assist physicians working with HSP patients by providing objective, automated gait parameters. To the best of the authors' knowledge, this is the first paper to investigate solutions for mobile gait analysis of patients affected by HSPs. Ultimately, automated, mobile gait analysis of HSP patients would allow ongoing and long term monitoring, providing useful insights into this orphan disease.