Noninvasive Early Respiratory Illness Diagnosis
Inexpensive and quick physical exam procedures such as auscultation (e.g. using a stethoscope), percussion, and tactile fremitus have enabled physicians to assess the health of the respiratory systems for centuries. Utilization of sensor arrays and ultrasound coupled with advanced signal processing techniques for feature extraction and classification of the respiratory sounds and vibrations resulted in a better understanding of lung acoustics under healthy and pathologic conditions.
In our lab, we develop computational models as well as signal processing and machine learning algorithms to characterize these signals and utilize them toward developing novel diagnosis methods for respiratory diseases.
Develop early diagnosis method for respiratory illnesses based on respiratory-induced sounds and vibrations measured noninvasively on the chest surface
Signal and image processing
Computational fluid dynamics and finite element modeling
Sensor array design