More than one third of patients suffering severe traumatic brain injury (TBI) exhibit dysarthria (Gilchrist & Wilkinson, 1979). Besides compromised articulation (Yorkston, Beukelman & Bell, 1988), two thirds of these patients are perceived to be dysphonic to some degree (Theodoros, Murdoch & Chenery, 1994). In the early stage post-injury, voice quality in dysarthric TBI speakers is perceived as mainly weak and breathy, whereas in later stages a harsh and strainedstrangled voice quality predominates (Theodoros & Murdoch, 1994; Vogel & von Cramon, 1982). Auditory-perceptual evaluation has been the major clinical approach in diagnosis of voice disorders. In addition, electroglottographic (Theodoros & Murdoch, 1994) and acoustic analyses (Hartmann & von Cramon, 1984) have been performed to determine the nature of the vocal fold vibratory movements as well as the frequency and severity of deviant laryngeal features. However, electroglottographic (EGG) measures yielded rather nonhomogenous profiles of abnormalities (Theodoros & Murdoch, 1994) and most acoustic measures were found to be sensitive to several perceptual dimensions such as roughness and breathiness and, hence, may be ambiguous with respect to the underlying pathophysiological mechanisms, i.e. irregularity of vocal fold oscillations and insufficiency of vocal fold closure (Michaelis, Fröhlich & Strube, 1998). Recently, Michaelis et al. (1998) succeeded in deriving two largely independent parameters from the acoustic signal that are related to distinct perceptual qualities as well as to different modes of vocal tract excitation in organic voice disorders. To test the clinical usefulness of the so-called Goettinger Hoarseness Diagram (GHD) within the domain of neurogenic voice disorders, this study first applied this procedure to dysphonia subsequent to severe TBI and, second, compared these
findings to auditory-perceptual ratings, another acoustic analysis program, MultiDimensional Voice Program (MDVP), and EGG measures.