An intelligent quantitative analysis software for videofluoroscopic swallowing study in patients with dysphagia

Area of Science:

• Medical Imaging
• Biomechanical Engineering
• Clinical Diagnostics

Background:

• Videofluoroscopic swallowing study (VFSS) is crucial for objective dysphagia diagnosis.
• Conventional VFSS quantitative analysis is time-consuming and labor-intensive.
• Existing automated methods require manual input and offer limited metrics.

Purpose of the Study:

• To develop an intelligent VFSS quantitative analysis tool for automated tracking.
• To provide comprehensive kinematic and temporal swallowing parameters.
• To assist clinicians in efficient and accurate dysphagia diagnosis.

Main Methods:

• Utilized a feature-based target tracking and contour extraction algorithm.
• Automated detection of hyoid bone displacement, UES opening, and pharyngeal contraction.
• Analyzed 82 VFSS samples (40 dysphagia patients, 42 healthy controls).

Main Results:

• Strong correlations (r=0.947-0.995) between automated and manual tracking for kinematic parameters.
• Low relative errors (2-5%) in automated parameter measurements.
• Demonstrated significantly reduced swallowing biomechanics in dysphagia patients.
• Reduced analysis time from ~1 hour to 3-4 minutes per case.

Conclusions:

• The developed software offers an efficient, user-friendly platform for VFSS quantitative analysis.
• Automated tracking and assessment enhance diagnostic accuracy and workflow efficiency.
• The tool streamlines dysphagia diagnosis for clinicians.