Figure 6 | Scientific Reports

Figure 6

From: Extraction of apex beat waveform from acoustic pulse wave by sound sensing system using stochastic resonance

Figure 6

Classification of Front CAB (0.5-BF) waveform. (a) The time waveforms for one second of ECG of five subjects. (b) The waveforms of Front CAB (0.5-BF) are classified into Square (S), Triangular (T), Cosine (C), Gauss (G) and Random (R) by their forms. The “A” wave of ACG (wave of atrial contraction: arrow head) and a similar waveform of ventricular systolic wave, which starts from the E point (*), are recognized. (c) PSD obtained by the frequency analysis of Front CAB (0.5-BF) revealed power spectra of the fundamental vibration and higher harmonics. PSD was divided into three patterns: Pattern (X) in which only the fundamental vibration is predominent, pattern (Y) in which the higher harmonics declined from the fundamental vibration naturally, and pattern (Z) in which multiple higher harmonics existed without diminishing from the fundamental vibration. (d) STFT analysis of the Front CAB (0.5-BF) waveforms was performed depending on two frequencies, 1.95 Hz for a power spectrum component of the A waveform (five lower panels), and 7.81 Hz for that of the E point (five upper panels). Results of STFT analysis were classified into five patterns of (I),(II), (III), (IV) and (V). Figure 6d shows STFT figures of the Front CAB (0.5-BF) time waveforms to readily conduct pattern matching by doubling the information amount on the power spectrum with two frequencies of 1.95 Hz and 7.81 Hz. They were classified into five patterns of (I), (II), (III), (IV) and (V). (e) Of the correlograms of Front CAB (0.5-BF), (i) which is a sine wave, (ii) which is a waveform with a notch on a sine wave, and (v) which is a waveform with two or more notches on a sine wave, are extracted. The remaining waveforms are divided into (iii), in which one side of the envelope curve is decayed, and (iv), in which both sides of the envelope curve are decayed.

Back to article page