Number of the records: 1
Discrimination of normal and abnormal heart sounds using probability assessment
- 1.
SYSNO ASEP 0474793 Document Type C - Proceedings Paper (int. conf.) R&D Document Type Conference Paper Title Discrimination of normal and abnormal heart sounds using probability assessment Author(s) Plešinger, Filip (UPT-D) RID, ORCID, SAI
Jurčo, Juraj (UPT-D) RID, SAI
Jurák, Pavel (UPT-D) RID, ORCID, SAI
Halámek, Josef (UPT-D) RID, ORCID, SAINumber of authors 4 Source Title Computing in Cardiology (CinC) 2016, 43. - Vencouver : Computing in Cardiology, 2016 - ISSN 2325-8861 - ISBN 978-1-5090-0896-4 Pages s. 801-804 Number of pages 4 s. Publication form Print - P Action Computing in Cardiology (CinC) 2016 Event date 11.09.2016 - 14.09.2016 VEvent location Vencouver Country CA - Canada Event type WRD Language eng - English Country CA - Canada Keywords heart ; feature extraction ; training ; histograms Subject RIV JA - Electronics ; Optoelectronics, Electrical Engineering OECD category Medical engineering R&D Projects GAP102/12/2034 GA ČR - Czech Science Foundation (CSF) LO1212 GA MŠMT - Ministry of Education, Youth and Sports (MEYS) ED0017/01/01 GA MŠMT - Ministry of Education, Youth and Sports (MEYS) Institutional support UPT-D - RVO:68081731 UT WOS 000405710400201 DOI 10.22489/CinC.2016.233-260 Annotation According to the “2016 Physionet/CinC Challenge”, we propose an automated method identifying normal or abnormal phonocardiogram recordings. Method: Invalid data segments are detected (saturation, blank and noise tests). The record is transformed into amplitude envelopes in five frequency bands. Systole duration and RR estimations are computed, 15-90 Hz amplitude envelope and systole/RR estimations are used for detection of the first and second heart sound (S1 and S2). Features from accumulated areas surrounding S1 and S2 as well as features from the whole recordings were extracted and used for training. During the training process, we collected probability and weight values of each feature in multiple ranges. For feature selection and optimization tasks, we developed C# application PROBAfind, able to generate the resultant Matlab code. Results: The method was trained with 3153 Physionet Challenge recordings (length 8-60 seconds, 6 databases). The results of the training set show the sensitivity, specificity and score of 0.93, 0.97 and 0.95, respectively. The method was evaluated on a hidden Challenge dataset with sensitivity and specificity of 0.77 and 0.91, respectively. These results led to an overall score of 0.84. Workplace Institute of Scientific Instruments Contact Martina Šillerová, sillerova@ISIBrno.Cz, Tel.: 541 514 178 Year of Publishing 2018
Number of the records: 1