Impaired swallowing (oropharyngeal dysphagia or OPD) causes nearly 150,000 annual hospitalizations and over 220,000 additional hospital days, and prolongs hospital lengths of stay by 40%. OPD risk is typically identified through subjective standard institutional screening (SIS) protocols and those identified through screening undergo gold standard imaging testing such as videofluoroscopy (VF). However, SIS methods over- or underestimate risk, and completely fail to identify patients with silent OPD who silently aspirate food into their lungs, raising their risk of pneumonia and other adverse events. Pre-emptive detection of silent or near- silent aspiration is essential. Our long-term goal is to develop an instrumental dysphagia screening approach based on high-resolution cervical auscultation (HRCA) to accurately predict OPD-related adverse events, and initiate more timely intervention measures to mitigate them. The overall objective here is to develop accurate, advanced data analysis approaches to translate HRCA signals to swallowing events observed in VF images. Our strong preliminary data has led us to our central hypothesis: HRCA coupled with advanced data analytics tools are powerful approaches to automate and improve existing dysphagia screening protocols. The rationale is that a reliable, robust early-warning instrumental OPD screening approach will reduce adverse events in patients with silent aspiration/dysphagia, shorten length of stay, reduce cost, and improve patient health. Guided by strong preliminary data, we will pursue the following three specific aims: (1) define HRCA signal signatures that characterize the entire continuum swallowing safety from unimpaired to severely impaired; (2) translate HRCA swallow signal signatures and equate them to validated measures of swallowing impairment; and (3) prospectively assess the effectiveness of our HRCA system in predicting clinically significant OPD and aspiration in a randomized, controlled trial. Under the first aim, we will collect HRCA swallow signatures from unimpaired people, and combine and analyze them along with our large database of swallows of people with OPD to characterize the entire range of swallowing function from unimpaired through severe OPD. Under the second aim, we will develop HRCA OPD severity cutoffs and match them to gold standard derived OPD impairment cutoffs to establish HRCA?s ability as a diagnostic surrogate that differentiates clinically significant OPD and aspiration from benign swallowing impairments. Under the third aim, we will test HRCA in a clinical setting by deploying HRCA with consenting patients, and comparing the accuracy of independent HRCA, independent SIS, and HRCA+SIS to VF data from all participants. The approach is innovative, as it will combine technology with clinical judgment to shift the OPD screening paradigm and fundamentally improve efforts to reduce morbidity and mortality caused by OPD. Our work is significant, because it will translate to an early-warning HRCA screening tool that will elevate the current standard of patient care by ensuring that patients with OPD are correctly identified before adverse events can occur.

Public Health Relevance

The proposed research is relevant to public health as hundreds of thousands of people annually suffer from compromised nutrition, hydration, and quality of life due to impaired swallowing and aspiration of swallowed food into the respiratory system, which increases the risk of death during hospitalization threefold, and adds 53,000 annual hospitalization days and nearly $550 million to US health care costs annually. The proposed research is relevant to the part of NIH?s mission that pertains to enhancing health, lengthening life and reducing illnesses, as we will develop an objective, precise instrumental dysphagia screening tool that that will fundamentally advance efforts to reduce morbidity and mortality caused by impaired swallowing.

National Institute of Health (NIH)
Eunice Kennedy Shriver National Institute of Child Health & Human Development (NICHD)
Research Project (R01)
Project #
Application #
Study Section
Biomedical Imaging Technology Study Section (BMIT)
Program Officer
Cruz, Theresa
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
University of Pittsburgh
Engineering (All Types)
Biomed Engr/Col Engr/Engr Sta
United States
Zip Code
Zhang, Zhenwei; Coyle, James L; Sejdi?, Ervin (2018) Automatic hyoid bone detection in fluoroscopic images using deep learning. Sci Rep 8:12310
Sejdi?, Ervin; Orovi?, Irena; Stankovi?, Srdjan (2018) Compressive sensing meets time-frequency: An overview of recent advances in time-frequency processing of sparse signals. Digit Signal Process 77:22-35
Sejdi?, Ervin; Orovi?, Irena; Stankovi?, Srdjan (2018) A software companion for compressively sensed time-frequency processing of sparse nonstationary signals. SoftwareX 8:9-10
Dudik, Joshua M; Kurosu, Atsuko; Coyle, James L et al. (2018) Dysphagia and its effects on swallowing sounds and vibrations in adults. Biomed Eng Online 17:69
Movahedi, Faezeh; Coyle, James L; Sejdic, Ervin (2018) Deep Belief Networks for Electroencephalography: A Review of Recent Contributions and Future Outlooks. IEEE J Biomed Health Inform 22:642-652
Dudik, Joshua M; Coyle, James L; El-Jaroudi, Amro et al. (2018) Deep Learning for Classification of Normal Swallows in Adults. Neurocomputing 285:1-9
Movahedi, Faezeh; Kurosu, Atsuko; Coyle, James L et al. (2017) Anatomical Directional Dissimilarities in Tri-axial Swallowing Accelerometry Signals. IEEE Trans Neural Syst Rehabil Eng 25:447-458
Jestrovi?, Iva; Coyle, James L; Sejdi?, Ervin (2017) A fast algorithm for vertex-frequency representations of signals on graphs. Signal Processing 131:483-491
Movahedi, Faezeh; Kurosu, Atsuko; Coyle, James L et al. (2017) A comparison between swallowing sounds and vibrations in patients with dysphagia. Comput Methods Programs Biomed 144:179-187
Jestrovi?, Iva; Coyle, James L; Sejdi?, Ervin (2017) Differences in brain networks during consecutive swallows detected using an optimized vertex-frequency algorithm. Neuroscience 344:113-123

Showing the most recent 10 out of 22 publications