Dynamic 3D Print of the Breathing Function

Abstract

Waveforms extracted via nasal thermistors are the most common signals used to study breathing function in sleep studies. In recent years, unobtrusive alternatives have been developed based on thermal imaging. Initially, the research aimed to produce a measurement on par with the clinical standard (the thermistor), but at a distance. Lately, there has been recognition that imaging is inherently multidimensional and can produce spatiotemporal and not just temporal signals - a development with significant diagnostic value. The extraction of 3D breathing information, however, has been based on inaccurate assumptions regarding the formation of the nasal thermal patterns sensed by the camera. The present paper corrects these assumptions, enabling the production of more accurate and complete multidimensional breathing signals.