Objective computer-based assessment of valleculae residue--is it useful?

Area of Science:

• Clinical Swallowing Assessment
• Medical Imaging Analysis
• Speech-Language Pathology

Background:

• Subjective rating scales for valleculae residue from fluoroscopic images have raised concerns regarding their reliability.
• Accurate assessment of valleculae residue is crucial for diagnosing and managing dysphagia.

Purpose of the Study:

• To evaluate the interrater agreement of a traditional subjective scale for valleculae residue.
• To compare the reliability of the traditional scale with a novel objective measurement scale, the valleculae residue ratio.

Main Methods:

• Five speech-language therapists assessed 100 still fluoroscopic images from 20 patients using a subjective visual grading scale (none, mild, moderate, severe).
• The same images were reassessed by the clinicians using Picture Archiving Communication System (PACS) measurement tools to calculate the valleculae residue ratio.
• A linear classifier was used to define cutoff points for the valleculae residue ratio scale, categorizing residue severity.

Main Results:

• The new valleculae residue ratio method demonstrated reliability comparable to the traditional subjective method.
• Interrater reliability: kappa = 0.73 for both methods.
• Intrarater reliability: kappa = 0.87 for the new method versus 0.85 for the traditional method.

Conclusions:

• The valleculae residue ratio is proposed as a new, quick, and reliable method for quantifying residue when PACS is available.
• This objective method offers a potentially more consistent alternative to subjective visual grading in clinical practice.
• Further testing is warranted to assess the impact of this method in settings with known poor inter- and intrarater reliability.