MV-MS-FETE: Multi-view multi-scale feature extractor and transformer encoder for stenosis recognition in echocardiograms

Danilo Avola, Irene Cannistraci, Marco Cascio, Luigi Cinque, Alessio Fagioli, Gian Luca Foresti, Emanuele Rodolà, Luciana Solito

Abstract: Highlights•Exploring the multi-view paradigm to predict aortic stenosis in echocardiogram images.•Designing a multi-view (PLAX/PSAX), multi-scale features, and transformer encoder model (MV-MS-FETE) to diagnose stenosis.•Presenting single-view and multi-view benchmarks on the recently released Tufts medical echocardiogram public dataset.•Setting a new multi-view state-of-the-art on stenosis recognition in terms of accuracy and F1-score on the Tufts dataset.