Characterization and Segmentation of Acoustic Swallowing Signals Collected Concurrently with Dual-axis Accelerometry

Dysphagia, a physiological impairment that involves difficulty swallowing, can arise due to a variety of differerent disease and injury processes. The current gold standard in assessment for dysphagia is the use of videofluoroscopy (VFS). Due to the need for radiation exposure in VFS, of the development of valid, noninvasive screening technologies is desirable. In this thesis, acoustics were explored as a potential means of screening for dysphagia. The foundation of this work requires pinpointing swallow segments in acoustic signals, which was done through the implementation of a novel acoustic swallow segmentation algorithm. The application of this algorithm to data from 44 healthy participants swallowing water samples resulted a sensitivity of 86%, 94% and 92% for detecting 10mL, 5mL and saliva swallows, respectively. Moreover, this work analyzed the suitability of different signal features for characterizing acoustic swallows and artifacts.M.H.Sc.2018-03-01 00:00:0