Monitoring Respiratory Motion With Wi-Fi CSI: Characterizing Performance and the BreatheSmart Algorithm

Abstract

Respiratory motion (i.e., motion pattern and rate) can provide valuable information for many medical situations. This information may help in the diagnosis of different health disorders and diseases. Wi-Fi-based respiratory monitoring schemes utilizing commercial off-the-shelf (COTS) devices can provide contactless, low-cost, simple, and scalable respiratory monitoring without requiring specialized hardware. Despite intense research efforts, an in-depth investigation on how to evaluate this type of technology is missing. We demonstrated and assessed the feasibility of monitoring and extracting human respiratory motion from Wi-Fi channel state information (CSI) data. This demonstration involves implementing an end-to-end system for a COTS-based hardware platform, control software, data acquisition, and a proposed processing algorithm. The processing algorithm is a novel deep-learning-based approach that exploits small changes in both CSI amplitude and phase information to learn high-level abstractions of breathing-induced chest movements and to reveal the unique characteristics of their difference. We also conducted extensive laboratory experiments demonstrating an assessment technique that can be replicated when quantifying the performance of similar systems. The results indicate that the proposed scheme can classify respiratory patterns and rates with an accuracy of 99.54% and 98.69%, respectively, in moderately degraded RF channels. Comprehensive data acquisition revealed the capability of the proposed system in detecting and classifying respiratory motions. Understanding the feasible limits and potential failure factors of Wi-Fi CSI-based respiratory monitoring scheme - and how to evaluate them - is an essential step toward the practical deployment of this technology. This study discusses ideas for further expansion of this technology.

Keywords: Channel state information; LSTM; MIMO-OFDM; Wi-Fi; deep learning; respiration monitoring; respiratory motion classification.

FIGURE 1.

Architecture of the proposed system.

FIGURE 2.

4D CSI tensor: A time series representative of CSI matrices for a MIMO-OFDM wireless network.

FIGURE 3.

FFT-based spectrum pre-processing on a CSI data stream of a normal breathing pattern with RR = 15 BPM.

FIGURE 4.

A high abstraction of LSTM network architecture used in this paper.

FIGURE 5.

Physical layout of the measurement setup inside the anechoic chamber.

FIGURE 6.

Block diagram of the measurement setup.

FIGURE 7.

Photograph of the measurement setup. The breathing manikin can be seen on a table in the foreground with its control box under a piece of the absorber to the right. The small shielded enclosure for the transmitting Wi-Fi device can be seen on the left side (dark-colored box) with the three antennas connected to a bulkhead panel on the enclosure. The receiving Wi-Fi device is in the center of the image; placed on a pedestal in the back of the chamber.

FIGURE 8.

Breathing rate classification for Att = 0dB and T = 60 sec. Each class corresponds to a respiratory rate, ranging from 3 BPM to 30 BPM. The misclassified predictions happen at (30, 12) and (15,27).

FIGURE 9.

Breathing pattern classification for Att = 0dB and T = 60 sec.

FIGURE 10.

Effect of attenuation on a complex CSI real and imaginary parts of a normal breathing with RR = 15 BPM.

FIGURE 11.

Heatmap confusion matrices. The rows and columns of the confusion matrix correspond to the predicted and true classes, respectively. Column-normalized elements display the percentages of correctly and incorrectly observation for each target class.

FIGURE 12.

Heatmap of the normalized confusion matrix –with more data compared to Fig. 11 (b)– for multi respiratory pattern classification when the transmitter is 72.76 meters away from the receiver.

FIGURE 13.

PDF estimation of attenuated Wi-Fi CSI data streams based on a normal Kernel function. Each line represents the PDF of Wi-Fi CSI amplitude values at a given attenuation value.

FIGURE 14.

Bidirectional LSTM architecture.

FIGURE 15.

LSTM cell.