Automatic Heart Sound Analysis for Cardiovascular Disease Assessment

Doctoral thesis English OPEN
Kumar, Dinesh; (2015)
  • Subject: doença valvular | classificação | análise tempo-frequência | wavelet decomposition | classification | valvular disease | som cardíaco | prosthetic valve | noise detection | dinâmica não linear | heart murmur | time-frequency analysis | decomposição de onduletas | heart sound | segmentation | nonlinear dynamics | segmentação | válvula prostética

Tese de doutoramento em Ciências e Tecnologias da Informação, apresentada ao Departamento de Engenharia Informática da Faculdade de Ciências e Tecnologia da Universidade de Coimbra Cardiovascular diseases (CVDs) are the most deadly diseases worldwide leaving behind d... View more
  • References (114)
    114 references, page 1 of 12

    Figure 2.8. Heart murmurs (one heart cycle demonstration), systolic in (a)-(f) and diastolic in (g)-(h). Systolic murmurs are: Aortic Regurgitation in (a), Aortic Stenosis (b), Mitral regurgitation (c), Pulmonary Stenosis (d), Systolic Ejection (e), and Ventricular Septal Defect in (f). And the diastolic murmurs are: Tricuspid Regurgitation in (g) and Mitral Stenosis in (h)...................................................................................................33

    Figure 2.9. Left figure is the time-frequency representation of a heart sound in mechanical valve (single tilted disk) at aortic position; and the right figure is time-frequency plot of the heart sound from biological valve at mitral position. ............................................................................................35

    Figure 2.10. The main cardiac intervals in the process of cardiac functioning (figure is regenerated from (Ahlstorm 2008)). .......................................................40

    Figure 3.1. Steps required in adaptive noise cancellation process in multi-channel signal approach. ...............................................................................................47

    Figure 3.2. Taped delay line adaptive line enhancement (ALE) filter model............49

    Figure 3.3. Steps required in segmentation for the sounds' components. .................54

    Figure 3.4. Segmentation of a heart sound using Liang's envelogram based method. The parameters c1 and c2 are the tolerance levels which are used to indentify diastolic and systolic intervals. ...................................................56

    Figure 3.5. Segmentation of the murmur portion in a heart sound with aortic stenosis. Heart murmurs are segmented based on the simplicity that is below the threshold. In the example shown simplicity is computed using embedding dimension m=4 and delay = 19 parameters (see Chapter 5 for the details of these parameters). .........................................59

    Figure 4.5. (a) Normal heart sound from native valve. (b) Spectrogram of (a). (c) Autocorrelation function. ..............................................................................82

    Figure 4.6. (a) Noisy heart sound from valve. (b) Spectrogram of (a). (c) Autocorrelation function. ..............................................................................83

  • Similar Research Results (3)
  • Metrics
    No metrics available
Share - Bookmark