The """"""""gold standard"""""""" for classification of motor speech disorders, known as the Mayo Clinic approach, is based on the work of Darley, Aronson and Brown (1969a, b;1975). By rating speech dimensions observed in patients with different neurological conditions, they identified five types of dysarthria, each possessing a unique clustering of perceptual features. Key to the classification system is that the underlying pathophysiology of each type of dysarthria is presumed responsible for the resulting clusters of perceptual features. This explanatory relationship between locus of damage and the perceptual features associated with a dysarthria has provided a convenient framework for clinical practice as well as research on motor speech disorders. However, the Mayo Clinic approach to differential diagnosis suffers significant limitations in that 1) there is substantial overlap of perceptual characteristics associated with the dysarthria classes, and 2) characteristics of a given dysarthria vary with severity. It is therefore difficult to perceptually classify speakers reliably without knowledge of underlying etiology (Fonville et al., 2008;Van der Graaff et al., 2009;Zyski and Weisiger, 1987). The reliability of classification problem extends beyond differential diagnosis to the extent that classification drives treatment decisions and research questions. An alternative taxonomical approach has been suggested by Weismer and Kim (2010) in which the goal is to identify a core set of common deficits in speakers with dysarthria. Identification of such similarities would allow the detection of differences that reliably distinguish different types of motor speech disorders irrespective of etiology. This, in turn, may inform intervention targets and induce a shift in research focus. The work proposed herein aims to make substantive contributions to the development of a perceptually-relevant taxonomy of dysarthria. Towards this end, segmental and/or suprasegmental acoustic correlates to perceptual dimensions associated with dysarthric speaker similarity will be investigated using cluster analysis and multidimensional scaling methods. Discriminant function analysis will be used to determine the potential of these dimensions for distinguishing among different dysarthria types (e.g., etiology-based and listener-derived categories). The results of this work are expected to establish a framework for a non-etiology based, perceptually-relevant taxonomy of dysarthric speech. This will permit the subsequent exploration of treatment targets and outcome measures in the clinical domain, and raise a new set of research questions in the domain of speech intelligibility.
Despite significant limitations, the current classification system for differential diagnosis of dysarthrias remains solidly the gold standard. This practice is an impediment to improving quality of care and expanding our knowledge base about the communication disorders associated with motor speech disorders. The current project aims to establish a framework for a non-etiology based, perceptually-relevant taxonomy of dysarthric speech that will permit subsequent exploration of treatment targets and outcome measures in the clinical domain, and raise a new set of research questions in the domain of speech intelligibility.
|Lansford, Kaitlin L; Liss, Julie M (2014) Vowel acoustics in dysarthria: speech disorder diagnosis and classification. J Speech Lang Hear Res 57:57-67|
|Lansford, Kaitlin L; Liss, Julie M (2014) Vowel acoustics in dysarthria: mapping to perception. J Speech Lang Hear Res 57:68-80|
|Lansford, Kaitlin L; Liss, Julie M; Norton, Rebecca E (2014) Free-classification of perceptually similar speakers with dysarthria. J Speech Lang Hear Res 57:2051-64|