Wearable Medical Device Ecosystems Through Machine Learning and Cloud Computing

Noncommunicable diseases including cardiovascular disease and diabetes have become the primary source of human mortality and disability in the 21st century, accounting for over 70% of deaths and 61% of disability-adjusted years worldwide. Medical infrastructure, which for decades sought to reduce the incidence and severity of communicable diseases, has proven insufficient in meeting the intensive and long-term monitoring needs of many noncommunicable disease (NCD) patient groups, particularly in low- and middleincome countries. However, recent trends in soft, wearable medical devices and cloud technologies offer a possible alternative to traditional in-person clinical monitoring. Soft sensors interfaced via a mobile device with the cloud could leverage this remote computational power to bring novel machine learning and signal processing to patients, displaying key health metrics such as heart rate and blood oxygen saturation in real-time and alerting to worrying trends. Using similar machine learning tools, this same cloud system could also provide long-term guidance to clinicians informed by the patient’s history, creating a pipeline between patients and the clinic. This is demonstrated here, wherein cloud-interfaced soft devices facilitate the remote monitoring of multiple, disparate NCD patient groups whose needs have proven difficult for contemporary clinical practices to accommodate, among them at-risk postpartum women and newborns with single ventricular heart disease. A mobile application, central cloud pipeline, and multiple soft sensors were designed to make a robust, highly scalable, remote monitoring ecosystem. In several clinical trials, this combination of novel soft device technology and delivery of cloud-based analysis to patient and clinician was shown successfully to detect and monitor disease progression, determine patient risk, and augment clinical decision-making for patient groups whose monitoring demands have proven intractable in standard clinical practice.

Level thesis:degree_level

Masters

Department dc:contributor.department

Mechanical Engineering

Grantor dc:publisher

Georgia Institute of Technology

Year dc:date.issued

2023

Author dc:creator

• Matthews, Jared

Advisor dc:contributor.advisor

• Yeo, Woon-Hong

Committee members dc:contributor.committeemember

• Lee, Seung-Woo
• Inan, Omer

• Wearable Device, Machine Learning, Cloud Computing

Language dc:language.iso

en_US

Handle dc:identifier.uri

https://hdl.handle.net/1853/75109

OAI identifier oai:identifier

oai:repository.gatech.edu:1853/75109