Gianna Cordova
Research Mentor(s): Alanson Sample
Mentor Department: Computer Science and Engineering
Authors: Gianna Cordova, Tze Tiong, Yasha Iravantchi, Alanson Sample
Session: Session 1 (9:00am – 9:50am)
Presentation Type: Poster 70
Abstract
The remote monitoring of medical conditions is critical for ensuring continuous patient care post-discharge. Current gaps in monitoring can lead to undetected complications, which could be prevented with timely interventions. This study aims to develop a privacy-sensitive application that leverages machine learning and Orange Pi technology to enable real-time remote health monitoring. Our specific focus is on enhancing remote monitoring by accurately identifying and labeling environmental sounds, thereby providing contextual data to healthcare providers. This project seeks to refine and implement an advanced system for home-based patient monitoring. Methodologies include using Orange Pi devices to collect environmental and physiological sound data from patients’ surroundings. These data are transmitted to an application for analysis, where machine learning algorithms are employed to recognize and label sounds accurately. To preserve privacy, data is anonymized and securely transmitted to Dropbox. Extensive testing of the application will assess functionality, data accuracy, and processing speed. Preliminary hypotheses suggest that this technology can accurately identify and label relevant home environment sounds, thus providing valuable insights for remote patient monitoring. Anticipated outcomes include improved monitoring capabilities that allow healthcare providers to intervene more effectively, ultimately enhancing patient care. This research highlights the potential for integrating sound recognition technology into broader telehealth systems, addressing the need for more effective remote health monitoring. By leveraging machine learning, this approach aims to improve healthcare for patients
