Toward Explainable Deep Learning in Healthcare through Transition Matrix and User-Friendly Features
Frontiers in Artificial Intelligence, Vol. 7, Article 1482141
Editorial summary
A transition matrix connects deep-network outputs with clinical features selected using guidelines and expert rules. ECG arrhythmia and cardiac MRI experiments compare these explanations with expert annotations. Agreement is measured separately for the two datasets.
Bibliographic reference
Use the publisher’s record for the citation format required by your journal or organisation.
Oleksander Barmak, Iurii Krak, Sergiy Yakovlev, Eduard Manziuk, Pavlo Radiuk, Vladislav Kuznetsov. "Toward Explainable Deep Learning in Healthcare through Transition Matrix and User-Friendly Features". Frontiers in Artificial Intelligence, Vol. 7, Article 1482141, 2024. https://doi.org/10.3389/frai.2024.1482141
Test a method against your business task.
Describe the decision your system needs to support and the data you have. We assess whether a method fits, define a prototype and agree how to measure its quality, response time and operating cost.