Low-complexity framework for movement classification using body-worn sensors

Abstract

We present a low-complexity framework for classifying elementary arm-movements (reach-retrieve, lift-cup-to-mouth, rotate-arm) using wrist-worn, inertial sensors. We propose that this methodology could be used as a clinical tool to assess rehabilitation progress in neurodegenerative pathologies tracking occurrence of specific movements performed by patients with their paretic arm. Movements performed in a controlled training-phase are processed to form unique clusters in a multi-dimensional feature-space. Subsequent movements performed in an uncontrolled testing-phase are associated to the proximal cluster using a minimum distance classifier (MDC). The framework involves performing the compute-intensive clustering on the training-dataset offline (Matlab) whereas the computation of selected features on the testing-dataset and the minimum distance (Euclidean) from pre-computed cluster centroids are done in hardware with an aim of low-power execution on sensor nodes.
The architecture for feature-extraction and MDC are realized using Coordinate Rotation Digital Computer based design which classifies a movement in (9n+31) clock cycles, n being number of data samples. The design synthesized in STMicroelectronics 130nm technology consumed 5.3 nW @50 HZ, besides being functionally verified upto 20 MHz, making it applicable for real-time high-speed operations. Our experimental results show that the system can recognize all three arm-movements with average accuracies of 86% and 72% for four healthy subjects using accelerometer and gyroscope data respectively, whereas for stroke survivors the average accuracies were 67% and 60%. The framework was further demonstrated as a FPGA-based real-time system, interfacing with a streaming sensor unit.

More information

Identifiers

Catalogue record

Export record

Altmetrics