|
Article Excerpt
Emerging data suggest that aprosodia may be amenable to behavioral treatment. This study investigated the use of acoustic analysis of speech to quantify response to two speech treatments previously judged to have an effect based on perceptual assessment in three participants with primarily expressive aprosodia. The mean and variability of fundamental frequency (F0) and intensity (INT) during production of sentences requiring use of four different emotional tones of speech (i.e., happy, angry, sad, or neutral) was calculated before and after two mechanism-based treatments for aprosodia (i.e., TX1 and TX2). Statistical differences in F0 mean and variability were primarily observed following TX1, whereas differences in INT mean and variability were principally revealed following TX2. Additionally, significant differences in these acoustic values were noted across almost all pairwise comparisons of emotional sentence types (i.e., angry vs. sad, happy vs. sad, neutral vs. sad, angry vs. neutral, and happy vs. neutral). These preliminary data suggest that perceptual improvements in aprosodia can be measured quantitatively using acoustic analysis of speech and provide additional support for previously described behavioral treatments for this disorder. These findings also support previous reports that suggest that different emotional tones of speech are associated with differences in the acoustic speech signal.
**********
Prosody can be defined as the suprasegmental features of speech conveyed perceptually by pitch, stress, and duration. These features have also been defined acoustically primarily as changes in fundamental frequency (F0), intensity (INT), and timing (Kent & Read, 2002; Kent, Weismer, Kent, Vorperian, & Duffy, 1999; Lehiste, 1970; Pell 1999). Prosody has been described as having both linguistic and affective components (Heilman, Leon, and Rosenbek, 2004; Pell, 1999). Linguistic prosody serves to clarify word and sentence types and affective (or emotional) prosody is used to express a speaker's emotions and attitudes (Myers, 1999). Traditionally, the left hemisphere has been assigned the major role in controlling linguistic prosody and the right hemisphere in controlling affective prosody, though this has been viewed by some as an oversimplification (Pell, 1999). Abnormalities in affective prosody have been described as expressive (Tucker, Watson, & Heilman, 1977) or receptive in nature (Heilman, Scholes, & Watson, 1975). Ross (1981; see also Ross & Monnot, 2008) suggested that affective prosodic deficits associated with right hemisphere lesions be called aprosodia and posited specific types that mirror the traditional aphasia types.
The effect of aprosodia on functional status has prompted growing interest in behavioral treatments, as reflected in several studies (Anderson, Beversdorf, Heilman, & Gonzalez-Rothi, 1999; Leon et al., 2005; Rosenbek et al., 2004; Stringer, 1996). Most recently, Rosenbek et al. (2006) administered two randomly ordered conceptual treatments, one imitative and one cognitive-linguistic, to 14 individuals with primarily expressive aprosodia after right-hemisphere stroke. The imitative program was motivated by the hypothesis that expressive aprosodia results from impaired motor programming/planning of the vocal elements that constitute emotional prosody (van der Merwe, 1997), whereas the cognitive-linguistic approach was motivated by the hypothesis that aprosodia results from a deficit in a modality specific nonverbal affect lexicon (Bowers, Bauer, & Heilman, 1993).
An ABAC single-subject design replicated across subjects was employed. Probes of treated and untreated emotions were administered during baseline and treatment phases, and these probe items were then perceptually judged by a trained rater blinded to the testing condition. These data were used to perform visual inspection and determine effect size. Visual inspection of the data was completed by three judges experienced in using this technique. For these 14 participants, visual analysis indicated a treatment effect for 18 out of 25 total treatments (three patients did not have the second treatment) and significant effect sizes (z [greater than or equal to] 2.00) were noted in 11 out of 25 total treatments.
Acoustic analysis of speech is being increasingly used in the study of affective prosody (Banse & Scherer, 1996; Erickson et al., 2006; Gandour. Larsen, Dechongkit, Ponglorpisit, & Khunadorn, 1995; Hammerschmidt & Jurgens, 2007; Kent & Rosenbek, 1982; Monnot, Orbello, Riccardo, Sikka, & Ross, 2003; Pell 1999; Pell & Baum, 1997; Scherer, 1986; Van Lancker & Sidtis, 1992). Right hemisphere lesions have been found to be associated with aberrant measures of F0, intensity, and timing at the sentence level (Gandour et al., 1995). Additionally, the notion that F0, particularly its variation, is "of primary importance in distinguishing affective meanings" (Pell & Baum, 1997, p. 208) has been supported by...
|
