Design of Analog Front End With Analog Features Extractions of EPHONOGRAM for Cardiac Timing Interval Estimation

Simultaneous recording of an electrocardiogram (ECG) and a phonocardiogram (PCG) is named as electro-phono-cardiogram (EPHONOGRAM). It helps in the analysis of the microvariations between the electrical and mechanical activity of the heart which is useful for diagnosing various cardiovascular diseases (CVDs) including myocardial contractility and coronary syndrome. In this article, analog front end (AFE) for simultaneous acquisition is presented. The ECG and PCG morphologies such as the R peak as well as <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${S}1$ </tex-math></inline-formula> and <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${S}2$ </tex-math></inline-formula> peaks are extracted in the analog domain to reduce the power consumption and computational complexity when compared to implementing the same on the digital signal processor (DSP). The cross-domain microvariations timing intervals such as the R–<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${S}1$ </tex-math></inline-formula> peak interval and the R–<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${S}2$ </tex-math></inline-formula> peak interval are obtained by exploiting the correlation between ECG and PCG timing features. Also, the fundamental heart sounds (<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${S}1$ </tex-math></inline-formula> and <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${S}2$ </tex-math></inline-formula>) from PCG are classified by correlating them with ECG features. The proposed circuit is simulated using UMC 0.18-<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\mu $ </tex-math></inline-formula>m CMOS technology. To analyze the performance of the AFE, performance matrices like input-referred noise (IRN), noise efficiency factor (NEF), power efficiency factor (PEF), and figure of merit (FoM) are used. To clinically validate the proposed system, the ECG and PCG signals are taken from the PhysioBank ATM database. For R peak detection, sensitivity of 100%, positive predictivity (PP) of 94.16%, and accuracy of 94.16% are achieved. Similarly, for <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${S}1$ </tex-math></inline-formula> and <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${S}2$ </tex-math></inline-formula> peaks detection, sensitivity of 97.81%, PP of 97.39%, and accuracy of 95.32% is obtained. The proposed concept is also verified in real time using the prototype built with discrete components and a Basys-3 FPGA board.