Automated class-based compression for real-time epileptic seizure detection

摘要: The emergence of next generation wireless networking technologies has motivated a paradigm shift in development viable mobile-Health applications for ubiquitous real-time healthcare monitoring. However, remote monitoring requires continuous sensing different biosignals and vital signs which results generating large volumes data that to be processed, recorded, transmitted. In this paper, we propose our vision the benefits leveraging edge computing enabling automated epileptic seizure detection. particular, an adaptive classification reduction technique reduces amount transmitted data, according class patients, while fast emergency notification patients with abnormality. Using such approach, patient aggregator can automatically reconfigures its compression threshold based on characteristics gathered maintaining required application distortion level. Our show excellent performance proposed scheme terms accuracy gain, as well advantages it exhibits respect state-of-the-art techniques.