Abstract-Cardiovascular diseases (CVDs) are currently the leading cause of deaths worldwide. The traditional auscultation is cost-effective and time-saving for the public to diagnose CVDs early. While many approaches for analysis of the heart sound (HS) signal from auscultation have been utilized successfully, few studies are focused on acoustic perspective to interpret the HS signal. This paper creatively proposes a multidimensional feature extraction technique based on timbre model to interpret HS, which stems from clinical diagnostic basis and medical knowledge. The extracted features have three dimensions, including spectral centroid (SC), log attack time (LT) and temporal centroid (TC). The simulation experiments indicate that the proposed method is promising in HS feature extraction and the later CVD diagnosis.Index Terms-Acoustics, feature extraction, Hilbert transform, spectral centroid, temporal centroid. I. INTRODUCTIONCardiovascular diseases (CVDs) have become a great threat to human' lives aggressively. A good way to prevent the death caused by CVDs is early discovery and interposition. Over the years, in spite of the advent of echocardiogram (ECHO), electrocardiogram (ECG), photoplethysmogram (PPG), sphygmogram (SPG), etc., the heart auscultation signal is still one of the most primary physiological signals. Physicians primarily analyze it thus to find signs of pathologic conditions, as it can provide clues to the diagnosis of many cardiac abnormalities including valvular heart disease, congestive heart failure and congenital heart lesions before requesting for ECHO, ECG, PPG, SPG, etc. Notably, in remote areas or less developed regions, auscultation may be the only means available. This leads the research on heart sound (HS) signal analysis and interpretation in order to provide a cost-effective and time-saving prognostic approach for the victims of CVDs.As for the analysis of HS, many approaches have been proposed in feature exaction in the literatures such as wavelet decomposition and reconstruction method . MFCC is based on the theory that human audition spaces are linearly at low frequency band and logarithmically at high frequency. As existing two inverse-transforms in MFCC, it encounters computation complexity.There exists another approach named timbre analysis from acoustic perspective, which hasn't been utilized and can avoid computation complexity. In acoustics, timbre is a significant attribute of three acoustic attributes, which embodies the texture of acoustic source. CVDs, as the pathological changes in the heart and the blood vessels, provide different timbre information. As a result, the timbre is well suitable for HS feature extractions. However, current timbre model analysis all aim at different music instrument recognition and little literature ever reported its exploration on HS feature extraction. In this paper, the timbre model based multidimensional feature extraction technique for HS analysis is creatively proposed. It is observed from simulations that the proposed method is ...

No abstract