Temporal Segmentation for Laryngeal High-Speed Videoendoscopy in Connected Speech.
Maryam Naghibolhosseini, Dimitar D Deliyski, Stephanie R C Zacharias and 2 others
PMID 28647431WHAT IT FOUND
An automated method matched a trained rater's manual segmentation of voicing and epiglottic obstruction in one healthy speaker's high-speed laryngeal video during connected speech.
It is a technical validation, not evidence of improved patient care.
Key findings
01Automatic segmentation agreed with visual segmentation for voiced segments and epiglottic obstructions, but some discrepancies were seen, for example around 4 seconds.
02Visual inspection of the entire recording, which had 116,543 frames, found no errors from the motion-window algorithm.
03The automatically computed frequency contour agreed with visual inspection of vibratory frequency changes in the high-speed video.
STILL TO COME
How it was doneWhat they foundWhat it means for SLPs
Read the rest of this summary
You get three full summaries a month, free, and we do not ask for a card. Search, the TL;DRs and your library stay unlimited either way.
What it does not show
Only one vocally normal participant was studied, so the method is not shown to work in patients with voice disorders. The authors state that more participants and different connected-speech samples are needed to address reliability and optimize the algorithm. The manual reference came from one trained rater, so agreement with other raters was not tested. Some discrepancies occurred between automatic and manual segmentation, including around 4 seconds. The study validates image processing, not a clinical diagnosis, treatment outcome, or patient care decision.
The easy way to misread this
Do not read this as proof that automated high-speed laryngeal video improves voice diagnosis or management. It validated a segmentation step in one healthy speaker, not in patients or clinical outcomes.