Implantable Cardioverter Defibrillators (ICDs) can help prevent cardiac arrest in patients with irregular heart rhythms, but only if they have suitable heart rhythms. Irregular heart rhythms, or arrhythmias, can lead to serious health complications, including stroke, heart failure, and sudden cardiac arrest, making accurate assessment and timely intervention crucial.
Evaluating a 24-hour ECG trace for suitability is time-consuming and complex, requiring expertise and significant manual effort. To address this, Prof. Alain Zemkoho’s group in the School of Mathematics has developed advanced machine learning techniques in collaboration with cardiologists from the University Hospital Southampton.
The SRSG took the initial research code and expanded it into a well-documented, modular library, ensuring greater accessibility and usability for medical professionals. Additionally, we developed a front-end interface for use in an exploratory trial of the technology in a clinical setting. This work streamlines the evaluation process, reduces workload and improves diagnostic accuracy, ultimately leading to better patient outcomes.