Simultaneous Monitoring of Canine Heart Rate and Respiratory Patterns During Scent Detection Tasks

Abstract

Man-made technological odor detection systems continue to improve but still cannot match the tracking efficiency, mobility, and selectivity in the presence of interfering odors achieved by detector dogs. The reliability of dogs as olfactory detectors does not depend solely on their performance but also on the handler’s skill in interpreting the behavior of the dog that signals the detection of the target odor. We present our efforts on a wireless wearable system combining electrocardiogram (ECG) and an electronic stethoscope for direct monitoring of cardiopulmonary events in dogs toward enabling cybernetic dog–machine interfaces. This is, to our knowledge, the first cyber-physical attempt to simultaneously record heart rate, heart rate variability, and continuous auscultation of respiratory behavior in a wearable form factor during scent-detection tasks. In this paper, we present: 1) a method to automate the detection of sniffing episodes based on power spectral density of the respiratory sounds; 2) proof-of-concept extraction and quantification of the ECG and respiratory sounds features that would enable the discovery of physiological patterns associated to scent-detection tasks; and 3) proof-of-concept correlation of such patterns with the presence/absence of target odors. These contributions pave the way for a novel real-time cybernetic olfactory detection monitoring system to provide decision support for handlers in the field in addition to enabling future computer-sniffing dog interfaces.