Joint Training of Hidden Markov Model and Neural Network for Heart Sound Segmentation

• Type: Article in International Conference Proceedings Book
• Date: 2021

Authors

Renna, F (Author) FCUP; Martins, ML (Author) Other; Coimbra, M (Author) FCUP

Conference proceedings International

Title: 2021 COMPUTING IN CARDIOLOGY (CINC)

Pages: 1-4

Conference on Computing in Cardiology (CinC)

Brno, CZECH REPUBLIC, SEP 12-15, 2021

Indexing

ISI Web of Knowledge - 1 Citation

Scopus - 2 Citations

Other information

Authenticus ID: P-00V-ZNN

DOI: 10.22489/cinc.2021.049

Abstract (EN): In this work, we propose a novel algorithm for heart sound segmentation. The proposed approach is based on the combination of two families of state-of-the-art solutions for such problem, hidden Markov models and deep neural networks, in a single training framework. The proposed approach is tested with heart sounds from the PhysioNet dataset and it is shown to achieve an average sensitivity of 93.9% and an average positive predictive value of 94.2% in detecting the boundaries of fundamental heart sounds.

Language: English

Type (Professor's evaluation): Scientific

No. of pages: 4