Monitoring and alerting system to determine muscle strength for fall risk assessment

Abstract

This report demonstrates the design of a healthcare monitoring and alerting system intended for the elderly using Electromyography (EMG) signals. The EMG signals are captured using a muscle sensor (Myoware muscle sensor) sampled using a microcontroller (Adafruit Feather HUZZAH) and relayed to a microcomputer (Raspberry Pi 4B+). Further python scripts and algorithms would then be run on the collated data in order to determine the elderly’s muscle strength in real-time, thereafter, alerting the elderly and/or their caregiver should their muscle state not deemed suitable for normal use. 2 prototypes are discussed in this paper. Both prototypes use the MQTT (Message Queuing Telemetry Transport) protocol to transmit data with the first using a RDBMS (Relational Database Management System) and the other using a Raw CSV (Comma-separated Values) file for data storage and processing. During this study, experiments were conducted to determine the relationship between muscle activations through real life situations that elderly face, such as having difficulty changing from a sitting position to a standing position. Research was also conducted to correlate grip strength with overall muscle usage. The two muscle groups used to conduct the tests were the Gastrocnemius Medialis and the Flexor Digitorum Profondus respectively. The results conclude that the prototype is able to detect muscle sensor anomalies with flexible detection mechanisms where required.