PNRR PE 14 RESTART: NETWIN, details at: https://www.fondazione-restart.it/projects/s3-netwin/
PNRR PE 14 RESTART: Net4Future, details at: https://www.fondazione-restart.it/projects/s14-net4future/
PRIN 2022: QT-SEED Quality-of-life Technological and Societal Exploitation of ECG Diagnostics
Details here: This project will address future scenarios where healthcare technologies allows health monitoring with wearable devices and triggering, if required, prompt interventions, especially in critical cases, e.g.,(the subject is alone, on board of a car, in crowded contexts). Health monitoring needs to focus on parameters that are easily measurable and contain meaningful information. Among these the ECG stands out. While it is true that ECG is one of the most studied physical signals, it is also true that it still carries a quantity of information that is still partially to be understood and exploited.
Few examples of still unexplored, but interesting, ECG issues follow: a) although a lot of research is done in the field of clinical interpretation both at the hospital level (12-lead ECG) and at the wearable ECG level (1 track), the analysis of ECG for health monitoring and stress (or other emotional states) management is less explored, despite its potentially high impact (also in association with pharmacological and behavioral digital therapies); b) there are still no works about collective ECGs, or about the analysis of the emotional state of groups of subjects in common conditions (e.g. in the cinema, in traffic), an application area with possible repercussions for the study of collective emotional effects; c) an innovative and promising approach to ECG analysis exploits the recent field of graph signal processing, representing the ECG as a signal defined on a mesh describing the heart morphology. In these domain, it is expected that novel vertex-time ECG features can be extracted, so as to empower signal analysis; d) ECG registrations do not have large data sizes and because of this there are not many works about ECG compression. ECG data size, however, becomes an issue if the signal is acquired with small devices, if acquisition is done 24/7 or collective data acquisition is performed. A model-based compression technique, leveraging, possibly, the ECG graph representation could be an useful tool both for ECG compression and automated analysis.
In this project the focus will be mainly on the above frontier aspects with activities linked to areas not yet explored.