Ulrich Eysholdt 1 , M. Döllinger 1 , J. Lohscheller 1

1 University Hospital, Dept. Phoniatrics, Erlangen

The clinical voice specialist, called phoniatrician, faces slight to severe cases of hoarseness. In general, hoarseness originates when the vocal fold vibrations develop irregularity and lose harmonicity. Following the trend of imaging in modern medicine, a method is needed to visualize irregular fold vibration. While the most wide-spread stroboscopy is very useful in normal or near-normal voice, i.e. harmonic, periodic vibration, it fails more and more with increasing vibration pathology. The reason is: in irregular vibration, stroboscopy violates the basic Shannon’s sampling theorema of signal processing. Highspeed recordings (HSR) in one or two dimensions overcome this problem. Combined with endoscopy, 2D-HSR are able to visualize irregular vibrations with real time resolution. However, HSRs are difficult to handle in a clinical environment. Even when replayed at slow motion HSR classification is difficult for the examiner, as motion is less obvious to human perception than static object features are. Under the name phonovibrogram (PVG) we developed a visualization procedure to make HSRs clinically more applicable. The PVG extracts the motion information from a complete HSR film and concentrates it into one single image at the cost of loss of the morphological anatomical information. The PVG shows a remarkable discriminatory power in any voice pathology. Evaluation criteria are (1) symmetry and (2) regularity of the vibration, which are readily identifiable in PVG. The PVG can be regarded as a fingerprint of the vibration and is suitable to machine classification. In a first approximation, automatic classification results can compete with those of experienced examiners. However, the factors causing irregularity cannot be derived from a PVG. They are studied in experiments and finite element model computations.