< Back to previous page

Project

Data-driven Models for Human Cardio-respiratory Event Monitoring

This PhD track resides in the 'Plug 'n Patch' personalized VLAIO ICON project that targets an innovative design methodology for medical patches that include customized sensors such as digital stethoscope (microphone) and an accelerometer. Innovative wearable sensors that provide high-quality data are required to monitor persons over a long-time span (e.g. weeks) in order to provide personalized health care. The candidate will mainly focus on automating the process of interpreting heart and respiratory sounds to support clinicians in improving the prognosis and disease monitoring of persons with cardio-pulmonary diseases. Efforts to apply machine learning algorithms to process and analyze medical images have demonstrated tremendous recent success, but similar achievements have not yet been realized for acoustic-based medical data. In this project we target a novel and multidisciplinary methodology that will produce a robust framework for interpreting cardio-respiratory sounds. When detected the results need to be visualized in a way that effectively conveys their meaning to the intended user. As a secondary focus, the PhD candidate will also work on the automated interpretation of acceleration signals that are of interest to keep track of the movement activity of a person (which will provide additional information next to that which is available in the heart and lung sounds). The research will be carried out in close interaction with doctors and biomedical scientists from the Mobile Health Unit (a structural collaboration on mHealth research between UHasselt, Ziekenhuis Oost-Limburg Genk and Jessa Hospital Hasselt).

Date:18 Sep 2020 →  Today
Keywords:cardio-pulmonary diseases, audio, signal processing, machine learning
Disciplines:Audio and speech computing, Biomedical signal processing, Machine learning and decision making, Signal processing
Project type:PhD project