توانبخشی مبتنی بر آواشناسی نام پریشی در زبان پریشی

This study investigated the effects of phonologic treatment for anomia in aphasia. We proposed that if treatment were directed at the level of the phonologic processor, opportunities for naming via a phonological route, as opposed to a strictly whole word route, would be enhanced, thereby improving naming. The participants, ten people with anomia and aphasia due to left hemisphere stroke, received 96 h of phoneme based treatment in 12 weeks. To learn if treatment improved naming, a single-subject, repeated probe design with replication was employed. The primary outcome measure was confrontation naming. Secondary outcome measures included phonologic production, nonword repetition and discourse production. Results suggest a positive treatment effect (confrontation naming), improvements in phonologic production and nonword repetition, and generalization to discourse production. When tested 3 months after the completion of treatment the effects appeared to be maintained.

One of the most common and debilitating features of aphasia is an impairment in ability to retrieve words, whether it involves naming seen objects, or producing nouns, verbs and other words conveying meaning in spontaneous propositional speech (Goodglass, 1993). The traditional treatment approach to this problem is to explicitly train individuals with aphasia in whole word naming (see Nickels, 2002 for extensive review). Controlled studies have shown that this approach may improve naming performance but generalization is typically very limited, that is, the knowledge gained by the patient tends to be limited to the words actually trained, and there is at best very modest improvement in naming performance with untrained words. This generalization may be limited mainly to words that are semantically related to those in the training corpus (Kiran and Thompson, 2003 and McNeil, 1997). The mechanisms underlying generalization are not well understood. Because generalization can be limited with “naming therapies”, there currently exists no viable means of training patients on the full corpus of words (perhaps several thousand) they are likely to need in daily life, except in the most determined and capable of subjects (Basso, 2003). Two approaches might be taken to solving this problem: (1) Develop cost effective means for providing training on several thousand words; and (2) develop alternative training methods. In this paper, we describe an alternative training method, training of phonemes and phonological sequence knowledge, that has the potential for broad generalization because knowledge of the full repertoire of phonemes and phonological sequence knowledge could potentially support production of all words. The mechanisms by which a purely phonological treatment could benefit anomia are implicit in a connectionist model of language function, discussed below.

Inter-class correlations assessing inter-rater reliability were 0.97 for confrontation naming, 0.99 for discourse word count, and 0.70 for discourse CIU. 3.2. Primary outcome results For the Object/Action Naming Test (confrontation naming), the average ES was 1.63 (SD 1.77) and graphs were visually judged to show evidence of generalization for 8/10 individuals (Fig. 2 and Fig. 3 and Table 2). 3.3. Secondary outcome results Refer to Table 3 (individual effect sizes and visual inspection) and Table 4 and Table 5 (individual raw scores in tabular display) for results. For phonologic production, the group average ES was 6.83 (SD 3.58) and graphs were visually inspected to show evidence of a treatment effect in 10/10 individuals. For nonword repetition, the average ES was 0.95 (SD 1.33) and graphs were visually inspected to show evidence of a treatment effect in 6/7 individuals. For discourse production, the average ES’s were 1.49 (SD 1.46) (word count) and 1.64 (SD 1.57) (CIU). Graphs were judged to show evidence of generalization for 4/6 individuals (word count) and 4/6 individuals (CIU). For the control probe, the average ES was 0.64 (SD 2.04) and graphs were judged to show evidence of a treatment effect in 2/8 individuals. Table 3.