Explainable AI helps solve tough cardiology cases, suggesting clinical adaptability by other specialties
Sifting the literature for real-world challenges thwarting adoption of clinical AI across medicine, a team of biomedical engineers and computer scientists has identified and fleshed out an exemplary use case.
It’s explainable AI-based analysis of electrocardiograms.
Senior author Yael Yaniv, PhD, and colleagues at the Israel Institute of Technology in Haifa settled on this application as one in which clinical AI has great potential within cardiology and possibly beyond.
In the course of completing the study, the authors demonstrated AI-based ECG analysis as both a cardiology-specific toolkit and an approach open to adaptation in other specialties.
They did so by challenging their model to take on—and complete—two tough tasks:
- detecting a wide array of heart arrhythmias of varying types, including unknown ones, and
- identifying core heart problems on apparently healthy segments of ECG readouts acquired from patients with intermittent arrhythmias.
Presenting their findings in Proceedings of the National Academy of Sciences, the researchers describe their work validating their methods on simulated arrhythmia screenings representing a diverse and populous caseload.
In the validation phase, Yaniv and team found their system:
1. visualizes the relative importance of each part of an ECG segment for the final model decision;
2. upholds specified statistical constraints on its out-of-sample performance and provides uncertainty estimation for its predictions;
3. handles inputs containing unknown rhythm types; and
4. handles data from unseen patients while also flagging cases in which the model’s outputs are not usable for a specific patient.
“Our results show that deep-learning systems can be robust, trustworthy, explainable and transparent while retaining the superior level of performance these algorithms are known for,” Yaniv and co-authors write.
What’s more, they state, the work represents “a significant step toward overcoming the limitations currently impeding the integration of AI into clinical practice in cardiology and medicine in general.”
Paper posted here (behind paywall).
Dave P. has worked in journalism, marketing and public relations for more than 30 years, frequently concentrating on hospitals, healthcare technology and Catholic communications. He has also specialized in fundraising communications, ghostwriting for CEOs of local, national and global charities, nonprofits and foundations.