An exploration of noncontact cardiopulmonary measurement using the smartphone in rescue relief events

This research has developed a noncontact cardiopulmonary detection method for natural disaster rescue relief events based on safe frequencies and directional antennae similar to those found in current smartphone-like devices. The vital sign sample datasets were collected in laboratory settings, which complied with ‘Ethical Standards’ applying to research involving human participants.
Equipment for correctly identifying living objects entrapped under heavy debris is generally purpose-built and costly; it must be operated by highly trained professionals and is not readily available in catastrophic events. In contrast, the low-cost alternative developed with this research represents an acceptable substitute for the commercially designed devices to detect heartbeat and respiration during rescue relief operations, thus reducing delays, device transit time, and cost, as well as eliminating the need for highly trained professionals.

The main research question seeks to determine if the safe industrial, scientific, and medical (ISM) frequency ranges currently deployed in smartphone-like devices are suitable for use as detectors’ medium. A further question investigates how supplementary equipment and approaches might enhance these devices to identify living objects under rubble or within a short distance of ‘Non-Line of Sight’ scenarios, i.e., directional antennae and Doppler effect radar detection techniques. In addition, the research tested the effectiveness of the proposed equipment ensemble in terms of the distance between the trapped living object and the Doppler radar hovering above the surface during data collection on heart rate and breathing beat. In addition to the collection of such significant wavelets emanating from the chest wall or breathing activity, the research also examined the viability of additional data collection from other areas of the human body. They included vascular-related vital signs generated by the groin, wrist, ankle, and cardiac neck. These datasets, once trained, were then used as a redundancy dataset to increase the efficacy of identifying the survivors of a disaster event.

Method: All datasets were collected from human objects - from body regions that are of significance for cardiopulmonary signal collection - from 360o degree views under simulated conditions of entrapment similar to those caused by natural or man-made disasters. The results were compared with traditional clinical contact-based vital sign methods to identify the living objects entrapped under debris. In addition, data were collected from individuals operating the sensor-equipped devices to identify the psychogenic tremors and behaviours likely to be experienced while engaged in rescue operations. Results have shown a clear relationship between the wavelength of pulmonary and blood vessel activities and the distance between the trapped human and the sensor in a range of conditions. This work significantly contributes to the existing research on timely rescue during disaster events utilising 2.4 GHz exiting in smartphone-like devices.

Therefore, if the application installed in smartphone-like devices has been adequately tested, verified, validated, and is functional, it could reduce the current average rescue time of 6.8 days (purpose-built equipment) to minutes and significantly increase the chance of saving human life. The research also proposed a theoretical model based on a simple anomaly algorithm for time-series applications based on an approach similar to that used for empirical data collection to form an anomaly algorithm.