UCSF

Objective

Although high-speed imaging (HSI) has been identified as a valuable tool in phonatory biomechanics research, to date, there have only been a selected number of reports investigating the clinical utility of HSI. We aim to elucidate the role of HSI in the diagnosis of the dysphonic patient.

Methods

The video files from 28 consecutive dysphonic patients with concurrently acquired videostroboscopy and HSI were retrospectively collected. Stroboscopy video files were edited to include vibratory motion only. Videos were then anonymously and randomly presented to four academic laryngologists. Experts were asked to assign a single best diagnosis for each video file. Assigned diagnoses were then compared with treatment diagnoses conferred based on medical history, phonatory evaluation, laryngeal examination, and response to treatment.

Results

Interrater analysis for the four laryngologists demonstrated significant and meaningful correlations for the diagnoses of polyps, cysts, nodules, and normal examination using stroboscopy (kappa > 0.40, P < 0.001). The experts demonstrated significant and meaningful correlations for the diagnoses of polyps, presbyphonia, and normal examination using HSI (kappa > 0.40, P < 0.001). Combined intrarater analysis performed by comparing single rater's diagnosis for single patient across both modalities resulted in poor correlation without statistical significance (kappa = 0.30, P = 0.07). Both stroboscopy- and HSI-assigned diagnoses matched the treatment diagnoses at equal predicted frequencies (32.3%), as demonstrated through multivariate logistic regression analysis (P < 0.001).

Conclusion

Overall, HSI did not improve the diagnostic accuracy above stroboscopy alone. Although specific laryngeal states such as presbyphonia may be better diagnosed with HSI, further studies are required to define HSI's precise role in the clinical setting.