A Framework for Prediction of Cramps during Activities of Daily Living in Elderly

A framework to predict cramps during Activities of Daily Living (ADL) to contribute towards assisted living and healthy aging of the constantly increasing elderly population, in the future of Internet of Things (IoT)-based interconnected living environments, for instance Smart Homes and Smart Cities is proposed in this paper. The framework leverages the immense potential at the intersection of IoT, Big Data, Human Computer Interaction, Artificial Intelligence and their interrelated disciplines. Leg cramps are known to be associated with all age groups but studies [1], [2] have shown that cramps are more frequent in older adults and have a multitude of risks and associated impacts, both temporary and permanent, which affect their quality of life and their abilities to perform ADL – which are essential for one’s sustenance. The challenge in this field is to predict cramps to minimize their effects on the performance of older adults during ADL, with an aim to improve their quality of life in the future of living environments. Addressing this challenge through development of a framework for Ambient Assisted Living (AAL) of elderly in Smart Homes serves as the main motivation for this work. To evaluate the efficacy of this proposed approach it has been tested on a dataset of ADL and the performance characteristics are discussed. A case study has also been performed where performance characteristics of various learning models are compared to deduce the best learning model for development of this framework. The results presented uphold the relevance and immense potential of this work for contributing towards AAL of the elderly population in Smart Homes through accurate prediction of cramps during ADL.