A Hybrid Machine Learning Method for Detecting Cardiac Ejection Murmurs
This paper presents a novel method for detecting cardiac ejection murmurs from other pathological and physiological heart murmurs in children. The proposed method combines a hybrid model and a time growing neural network for an improved detection even in
- PDF / 883,723 Bytes
- 4 Pages / 595.276 x 790.866 pts Page_size
- 92 Downloads / 188 Views
Department of Innovation, Design and Technology, Mälardalen University, Västerås, Sweden 2 CAPIS Biomedical Research and Department Center, Mons, Belgium 3 Department of Biomedical Engineering, Linköping University, Sweden
Abstract— This paper presents a novel method for detecting cardiac ejection murmurs from other pathological and physiological heart murmurs in children. The proposed method combines a hybrid model and a time growing neural network for an improved detection even in mild condition. Children with aortic stenosis and pulmonary stenosis comprised the patient category against the reference category containing mitral regurgitation, ventricular septal defect, innocent murmur and normal (no murmur) conditions. In total, 120 referrals to a children University hospital participated to the study after giving their informed consent. Confidence interval of the accuracy, sensitivity and specificity is estimated to be 87.2% ̶ 88.8%, 83.4% ̶ 86.9% and 88.3% ̶ 90.0%, respectively. Keywords— Intelligent phonocardiogram, heart sounds, machine learning, ejection murmurs.
I. INTRODUCTION
Cardiovascular disease is still the main factor of human mortality. In many heart defects, the mortality rate can be substantially decreased by timely screening and appropriate disease management [1,2]. Cardiac auscultation is considered as an important technique in clinical investigation. In this technique, assessment is based on the subjective decision of the listener whose expertise, experience and auditory capability play important roles in the accuracy of the result. Studies show that screening accuracy of the conventional auscultation is, yet insufficient, that can be improved either by training or the use of a computer assisted tool [3-5]. A heart produces two basic sounds in each cycle: the first heart sound (S1) and the second heart sound (S2). In addition to the basic sounds, a heart might initiate extra sounds, either due to pathological conditions or to an unusual physiological activity. Heart murmur is the dominant group of the extra sounds that may be heard either as a physiological murmur in 70% of the healthy children or as a pathological murmur in more than 80% of the pediatric heart pathologies [6,7]. Art of the cardiac auscultation is to properly perceive and to correctly classify the sounds emanating from heart. A large number of the pediatric referrals to hospitals for cardiac investigations have normal hearts while a number of the patients with the pathological signs are overlooked due to the complexity of the conventional auscultation [3]. © Springer Nature Singapore Pte Ltd. 2018 H. Eskola et al. (eds.), EMBEC & NBC 2017, IFMBE Proceedings 65, DOI: 10.1007/978-981-10-5122-7_197
Ejection murmur (EM) is a pathological sign of a stenosis in a semilunar valve (aortic or pulmonary valve), caused by the blood turbulence through the obstructed valve [6]. EM is heard right after the opening of the obstructed semilunar valve when the blood ejects into the arteries. Clinical manifestation of EM in children, is a diamo
Data Loading...
