Machine Learning Model Based on Transthoracic Bioimpedance and Heart Rate Variability for Lung Fluid Accumulation Detection: Prospective Clinical Study.

BACKGROUND: Accumulation of excess body fluid and autonomic dysregulation are clinically important characteristics of acute decompensated heart failure. We hypothesized that transthoracic bioimpedance, a noninvasive, simple method for measuring fluid retention in lungs, and heart rate variability, an assessment of autonomic function, can be used for detection of fluid accumulation in patients with acute decompensated heart failure. OBJECTIVE: We aimed to evaluate the performance of transthoracic bioimpedance and heart rate variability parameters obtained using a fluid accumulation vest with carbon black-polydimethylsiloxane dry electrodes in a prospective clinical study (System for Heart Failure Identification Using an External Lung Fluid Device; SHIELD). METHODS: We computed 15 parameters: 8 were calculated from the model to fit Cole-Cole plots from transthoracic bioimpedance measurements (extracellular, intracellular, intracellular-extracellular difference, and intracellular-extracellular parallel circuit resistances as well as fitting error, resonance frequency, tissue heterogeneity, and cellular membrane capacitance), and 7 were based on linear (mean heart rate, low-frequency components of heart rate variability, high-frequency components of heart rate variability, normalized low-frequency components of heart rate variability, normalized high-frequency components of heart rate variability) and nonlinear (principal dynamic mode index of sympathetic function, and principal dynamic mode index of parasympathetic function) analysis of heart rate variability. We compared the values of these parameters between 3 participant data sets: control (n=32, patients who did not have heart failure), baseline (n=23, patients with acute decompensated heart failure taken at the time of admittance to the hospital), and discharge (n=17, patients with acute decompensated heart failure taken at the time of discharge from hospital). We used several machine learning approaches to classify participants with fluid accumulation (baseline) and without fluid accumulation (control and discharge), termed with fluid and without fluid groups, respectively. RESULTS: Among the 15 parameters, 3 transthoracic bioimpedance (extracellular resistance, R0; difference in extracellular-intracellular resistance, R0 - R∞, and tissue heterogeneity, α) and 3 heart rate variability (high-frequency, normalized low-frequency, and normalized high-frequency components) parameters were found to be the most discriminatory between groups (patients with and patients without heart failure). R0 and R0 - R∞ had significantly lower values for patients with heart failure than for those without heart failure (R0: P=.006; R0 - R∞: P=.001), indicating that a higher volume of fluids accumulated in the lungs of patients with heart failure. A cubic support vector machine model using the 5 parameters achieved an accuracy of 92% for with fluid and without fluid group classification. The transthoracic bioimpedance parameters were related to intra- and extracellular fluid, whereas the heart rate variability parameters were mostly related to sympathetic activation. CONCLUSIONS: This is useful, for instance, for an in-home diagnostic wearable to detect fluid accumulation. Results suggest that fluid accumulation, and subsequently acute decompensated heart failure detection, could be performed using transthoracic bioimpedance and heart rate variability measurements acquired with a wearable vest.

Increased Conductivity and Reduced Settling Time of Carbon-Based Electrodes By Addition of Sea Salt for Wearable Application.

A carbon-based dry electrode is designed to measure bio-potential from skin surface without hydrogel. Consequently, unlike Ag/AgCl electrodes, the carbon-based electrodes require some settling time before a high-fidelity signal is obtained due to the process for impedance matching among skin surface, electrode and amplifiers in biometric system. Besides, especially, when electrocardiogram (ECG) is measured at some distance away from the chest using carbon-based electrodes for wearable application, the settling time could be a critical concern for immediate data collection due to the smaller bio-potential and bigger motion artifact noises. The settling time was defined as the time it takes for the carbon-based electrodes to have the same impedance as that of Ag/AgCl electrodes at a particular frequency (< 1 kHz) for bio-signals. In this study, we investigated the characteristics of the skin contact impedance as a function of time using carbon-based electrodes with and without sea salt and different thickness. Specifically, sea salt was added to the carbon black (SCB)/polydimethlysiloxane (PDMS) electrode to examine the level of enhanced conductivity and reduction of settling time. We used SCB/PDMS and CB/PDMS electrodes with thickness of 1.0 mm and 1.5 mm, examined their electrode and skin contact impedance values and compared them to Ag/AgCl electrodes. We collected impedance data from seven subjects using both SCB and CB/PDMS electrodes every 10 minutes for 50 minutes. A SCB/PDMS electrode showed lower impedance than a CB/PDMS electrode, and for both types of electrodes, higher thickness resulted in lower impedance. The same results were found for skin contact impedance. The settling times of the SCB/PDMS electrodes were found to be $20 \pm 10$ minutes and $40 \pm 10$ minutes for widths of 1.0 mm and 1.5 mm, respectively. The settling time for CB/PDMS without sea salt resulted in significantly higher settling time (> 50 minutes) when compared to SCB/PDMS electrodes. In summary, when carbon-based electrodes are used to measure bio-signals from skin surface for wearable application, its settling time can be partially offset by adding sea salt to CB/PDMS electrode and by making it thinner.

Analysis of Consistency of Transthoracic Bioimpedance Measurements Acquired with Dry Carbon Black PDMS Electrodes, Adhesive Electrodes, and Wet Textile Electrodes.

The detection of intrathoracic volume retention could be crucial to the early detection of decompensated heart failure (HF). Transthoracic Bioimpedance (TBI) measurement is an indirect, promising approach to assessing intrathoracic fluid volume. Gel-based adhesive electrodes can produce skin irritation, as the patient needs to place them daily in the same spots. Textile electrodes can reduce skin irritation; however, they inconveniently require wetting before each use and provide poor adherence to the skin. Previously, we developed waterproof reusable dry carbon black polydimethylsiloxane (CB/PDMS) electrodes that exhibited a good response to motion artifacts. We examined whether these CB/PDMS electrodes were suitable sensing components to be embedded into a monitoring vest for measuring TBI and the electrocardiogram (ECG). We recruited N = 20 subjects to collect TBI and ECG data. The TBI parameters were different between the various types of electrodes. Inter-subject variability for copper-mesh CB/PDMS electrodes and Ag/AgCl electrodes was lower compared to textile electrodes, and the intra-subject variability was similar between the copper-mesh CB/PDMS and Ag/AgCl. We concluded that the copper mesh CB/PDMS (CM/CB/PDMS) electrodes are a suitable alternative for textile electrodes for TBI measurements, but with the benefit of better skin adherence and without the requirement of wetting the electrodes, which can often be forgotten by the stressed HF subjects.

Feasibility Testing of Hydrophobic Carbon Electrodes for Acquisition of Underwater Surface Electromyography Data.

Underwater surface electromyography (sEMG) signals are especially of interest for rehabilitation and sports medicine applications. Silver/silver chloride (Ag/AgCl) hydrogel electrodes, although the gold standard for sEMG data collection, require waterproofing for underwater applications. Having to apply waterproof tape over electrodes impedes the deployment of sEMG in immersed conditions. As a better alternative for underwater applications, we have developed carbon black/polydimethylsiloxane (CB/PDMS) electrodes for collecting sEMG signals under water. We recruited twenty subjects to collect simultaneous recordings of sEMG signals using Ag/AgCl and CB/PDMS electrodes on biceps brachii, triceps brachii, and tibial anterior muscles. The Ag/AgCL electrodes were covered in waterproof tape, and the CB/PDMS electrodes were not. We found no differences in sEMG signal amplitudes between both sensors, for the three muscles. Moderate mean correlation between Ag/AgCl and CB/PDMS electrodes was found on the linear envelopes (≥ 0.7); correlation was higher for power spectral densities (≥ 0.84). Ag/AgCl electrodes performed better in response to noise, whilst the CB/PDMS electrodes were more sensitive to myoelectric activity in triceps and tibialis, and exhibited better response to motion artifacts in the measurements on the triceps and tibialis. Results suggest that sEMG signal collection is possible under water using CB/PDMS electrodes without requiring any waterproof or adhesive tape.