Exploring Wearable Systems for Seismocardiographic Information Extraction: Advancements in Utilization and Implementation

Seismocardiography (SCG), a technique measuring vibrations induced by the heart against the thoracic cage, provides valuable insights into cardiac function. It surpasses other clinical signals like ECG and PPG, offering a non-invasive assessment of cardiac mechanics, including contractility, valve movements, and key hemodynamic parameters. Phonocardiography (PCG) allows listening to internal sounds, but SCG, operating at lower frequencies (up to 40 Hz), extracts richer details. It introduces an automated approach, and it is more robust to ambient noise, holding potential for continuous remote monitoring. The evolution of SCG research spans physiological understanding, automated signal analysis, advanced wearables, exploration of cardiac indices, and clini- cal validation. These trajectories aim to harness the technique’s potential in reshaping clinical practices. My doctoral research addresses challenges in SCG advancement, focusing on sensor placement optimization and artifact removal during daily activities. Two wearable technologies were explored: a Fiber Optic-based Device for ambulatory settings and an Electronic Tattoo for continuous monitoring. Chapters are organized to delve into cardiac physiology and illustrate the systems explored to address challenges in wearable seismocardiography, namely optical wearable systems and electronic tattoos (i.e., e-tattoos). The unique wavelength division multiplexing (WDM) property of FBG sensors paved the way for the development of the first wearable system based on FBG sensors in an array configuration. This system enables the simultaneous acquisition of SCG signals from various points on the chest, allowing for a comprehensive analysis to determine the optimal position for signal recording. The proposed FBG-based system, following an extensive experimental campaign, demonstrated exceptional performance in Heart Rate (HR) monitoring near the xiphoid process. On an optimized version of the initially proposed FBG-based system, this location exhibited remarkable repeatability and minimal inter-subject variability, achieved through the implementation of a dedicated algorithm based on graph theory. This algorithm facilitates the evaluation of waveform similarity in SCG by extracting graphs from signals and calculating a measure of similarity between them. Notably, the xiphoid process outperforms other promising positions, such as the aortic valve, particularly in terms of signal waveform repeatability. Transitioning to the challenge of motion artifact removal, various sources of interference in SCG signals are identified, including general body movements, breathing, muscle tremors, bed or recording surface movements, improper sensor placement, and contact with external objects. Among these, body movement artifacts during daily activities emerge as the most challenging to remove. The study focuses on the implementation and optimization of an e-tattoo, specifically from the Lu Research Group at The University of Texas at Austin, to enable prolonged and preventive monitoring of cardiac function. This electronic tattoo was designed for simultaneous ECG and SCG monitoring. Here, the Normalized Least Mean Squares (NLMS) filter was proposed and demonstrated successful for motion artifact removal, allowing a prolonged monitoring of cardiac parameters during both rest and physical exertion. This study contributes significantly to the literature on e-tattoos, emphasizing artifact removal during motion for extracting clinically relevant parameters like Pre-Ejection Period (PEP) and Left Ventricular Ejection Time (LVET). The detailed explanation of the adaptive filtering technique, the device specifications, and the comprehensive data analysis and results discussion provide a thorough overview of the proposed technology. Future plans involve extensive testing for predictive capabilities on Stroke Volume (SV) and trials on patients with Cardiovascular Diseases (CVDs). The ultimate goal is to offer these wearable devices for continuous monitoring inside and outside the hospital setting, serving as a preventative tool for early diagnosis and personalized intervention in CVDs.