This shows you the differences between two versions of the page.

Link to this comparison view

simpler_faster_and_a_more_accurate_data_reduction_protocol_from_hsdi [2014/03/18 20:32] (current)
bziolko created
Line 1: Line 1:
 +Yuling Yan, Krzysztof Izdebski (USA)
 +**Simpler, faster and a more accurate data reduction protocol from HSDI**
 +High-speed digital imaging (HSDI) of the larynx can provide important information on the vocal fold kinematics. This information is key to understanding the mechanism of phonation and assisting clinical assessment of voice disorders. Tracking bilateral vocal fold vibrations is an important step for clinically relevant analysis and subsequent characterization of vocal fold vibrations, which are known to correlate with voice quality and laryngeal health condition. We introduce a new approach that combines the use of level-set segmentation and motion cue to exact vocal fold displacements. This approach is applied to videokymogram (VKG) that is reconstructed from a sequence of laryngeal images captured by the HSDI device. Further, motion cue is used to achieve effective and robust level-set based segmentation results. In particular, we construct a so-called STD (standard deviation) image by mapping on a pixel-by-pixel basis, the STD measure of the temporal intensity dispersion based upon initial HSDI sequence of images. The derived STD image maps the extent of vocal fold motion, and followed by simple threshold operation, a region of interest (ROI) that encloses maximum glottis excursion can be identified. The ROI is then used to restrict the operation of the level-set segmentation on VKG. The performance of the proposed approach is evaluated using clinical datasets representing both normal and pathological voice conditions.
Copyright © XXII PVC Organizing Committee 2013. All Rights Reserved.