Noninvasive detection of invertebrate respiratory dynamics via high-speed infra-red trachea imaging and analysis

Abstract: Respiration is a key physiological process, and tightly coupled with other physiological systems such as the neural system. Insects are emerging as critical model systems. Yet there is a lack of tools for insect respiration study, due to smaller size of the invertebrate respiratory system. We developed a system that employs high-speed infra-red imaging to record in insect movement of the trachea, the structure used for gas exchange. By image and signal processing with machine learning, we were able to detect tracheal motion signals that is consistent with manual obtained ground truth. Trachea motion detection provides key information on respiratory dynamics of insects, providing a novel research method for neuroscience and physiology, with potential applications in agriculture and forestry.