Wearable bioimpedance for continuous and context-aware clinical monitoring

Development of a Tetherless Bioimpedance Device That Uses Morphologic Changes to Predict Blood Flow Restrictions Mimicking Peripheral Artery Disease Progression

A tetherless multi-targeted bioimpedance device was designed, modeled, built, and tested for measuring arterial pulse and, using morphological analysis, its potential for monitoring blood flow restrictions that mimic Peripheral Artery Disease (PAD) was assessed across multiple peripheral arteries. Specifically, we first developed a small form factor, tetherless, bioimpedance device, based on high-frequency structure simulator (HFSS) simulations. After designing and building the device we then tested it in vivo on human subjects on multiple arteries and found that we did not need to modify the gain on the device compared to the bench top system. Further, it was found that changes in the morphology of the bioimpedance signal over time, depicted through the ratio of the first and second harmonic in the signal frequency, could be used to predict blood flow restrictions that mimic peripheral artery disease (PAD). The HFSS simulations helped guide the modulation frequency selection and the placement of the bioimpedance electrodes. We built the device and compared it to two commercially available bioimpedance devices and it was shown to demonstrate a distinct advantage in its multi-target capability, enabling more accurate pulse measurements from different arteries without the need for tuning the circuit for each artery. Comparing the ratio of the 1st and 2nd harmonics as a function of the blood flow restriction, the two commercial devices showed a maximum error across arteries of between 22% and 27% depending on the measurement location, whereas our system consistently displayed a stable value of just below 4%. With this system, there is the potential for comprehensive and personalized medical examinations for PAD at the point of care (POC).

A review of bio-impedance devices

Bio-impedance measurement analysis primarily refers to a safe and a non-invasive technique to analyze the electrical changes in living tissues on the application of low-value alternating current. It finds applications both in the biomedical and the agricultural fields. This paper concisely reviews the origin and measurement approaches for concepts and fundamentals of bio-impedance followed by a critical review on bio-impedance portable devices with main emphasis on the embedded system approach which is in demand due to its miniature size and present lifestyle preference of monitoring health in real time. The paper also provides a comprehensive review of various bio-impedance circuits with emphasis on the measurement and calibration techniques.

Towards Estimation of Tidal Volume and Respiratory Timings via Wearable-Patch-Based Impedance Pneumography in Ambulatory Settings

OBJECTIVE

Evaluating convenient, wearable multi-frequency impedance pneumography (IP)-based respiratory monitoring in ambulatory persons with novel electrode positioning.

METHODS

A wearable multi-frequency IP system was utilized to estimate tidal volume (TV) and respiratory timings in 14 healthy subjects. A 5.1 cm × 5.1 cm tetrapolar electrode array, affixed to the sternum, and a conventional thoracic electrode configuration were employed to measure the respective IP signals, patch and thoracic IP. Data collected during static postures-sitting and supine-and activities-walking and stair-stepping-were evaluated against a simultaneously-obtained spirometer (SP) volume signal.

RESULTS

Across all measurements, estimated TV obtained from the patch and thoracic IP maintained a Pearson correlation coefficient (r) of 0.93 ± 0.05 and 0.95 ± 0.05 to the ground truth TV, respectively, with an associated root-mean-square error (RMSE) of 0.177 L and 0.129 L, respectively. Average respiration rates (RRs) were extracted from 30-second segments with mean-absolute-percentage errors (MAPEs) of 0.93% and 0.74% for patch and thoracic IP, respectively. Likewise, average inspiratory and expiratory timings were identified with MAPEs less than 6% and 4.5% for patch and thoracic IP, respectively.

CONCLUSION

We demonstrated that patch IP performs comparably to traditional, cumbersome IP configurations. We also present for the first time, to the best of our knowledge, that IP can robustly estimate breath-by-breath TV and respiratory timings during ambulation.

SIGNIFICANCE

This work represents a notable step towards pervasive wearable ambulatory respiratory monitoring via the fusion of a compact chest-worn form factor and multi-frequency IP that can be readily adapted for holistic cardiopulmonary monitoring.