This aim of the project seeks to discriminate normal from abnormal airway protection and kinematic functions noninvasively via machine-learning analysis of Aspirometer/HRCA (high resolution cervical auscultation) signals, with similar accuracy as human judgment of VF. Hypothesis: Advanced data analytics can detect pathological airway protection in HRCA signal signatures with 90% accuracy when compared to a human expert's airway protection ratings from VF images. Analytical algorithms that can learn from data (e.g., Bayes' learning) will be used to infer about the continuum of abnormal airway protection during swallowing.
