With current advances in contact-free technologies, new ways to monitor patients with highly contagious diseases should be developed to ensure safe recovery. The majority of the self-quarantined patients in the United States have access to the Internet using WiFi connections. %WiFi is wireless networking technology that uses radio waves to provide wireless high-speed Internet and network connections.
In this project, we propose to create a novel system and framework to monitor patients with highly contagious diseases using contact free WiFi signals, complemented with noninvasive thermal information for inferring patient’s improvements and deterioration during the self-quarantine period. The major advantage of this project is that patients can use their own WiFi connection and afford a simple and affordable way to transmit information directly to their healthcare provider.
For effective RR estimation, we propose a high-resolution spectrogram based CSI approach for the Wi-COVID framework. The proposed approach combines advanced signal pre-processing techniques to effectively extract the target signal component, and the high resolution spectrogram to obtain an accurate and dynamic RR estimation.
The CSI signal processing consists of the following steps:
- Obtaining CSI signal magnitude;
- Removing outliers by the Hampel identifier, and using a band-pass filter with cut-off frequencies of 0.2 Hz and 0.4 Hz to suppress the noises;
- Using PCA to extract respiration component, and then applying high resolution spectrogram to extract the instantaneous frequency of the respiration component to estimate RR;
- Calculating the shortness of breath indicators in both time and frequency domains to evaluate the COVID-10 symptoms.
The sensing layer of Wi-COVID framework is composed by a off-the-shelf WiFi device and a Raspberry Pi that acts like a WiFi Access Point (AP). We utilize bcm43455c0 WiFi chip used in Raspberry Pi 4 for extracting CSI of OFDM-modulated Wi-Fi frames (802.11n) on a per frame basis with up to 80 MHz bandwidth.
Publications
[1] Li, Fangyu, Maria Valero, Hossain Shahriar, Rumi Ahmed Khan, and Sheikh Iqbal Ahamed. “Wi-COVID: A COVID-19 symptom detection and patient monitoring framework using WiFi.” Smart Health 19 (2020): 100147.