Temporal acuity rehabilitation and training

Recently, many independent research groups have suggested that temporal acuity training or rehabilitation could ameliorate the language or reading difficulties of SLI or dyslexic participants.

4.5.2.1 Specific language impairment

Merzenich, Tallal, and their colleagues (Merzenich et al., 1996) have shown that both temporal acuity and language comprehen-sion of children with SLI can be somewhat ameliorated by train-ing. In their first experiment (Merzenich et al., 1996) two tasks were trained with feedback (19-28 20-min sessions over the time period of 4 weeks): the order judgment of two upward/down-ward gliding stimuli (varying ISI and FM frequency) and a task in which the participant chose the predefined CV-target in a two alternative forced choice (2AFC) setting (the duration of the con-sonants, magnitude of consonant element amplification, and ISI between pairs were varied). The results in a pre/post-test Tallal Repetition Test (varying ISI and tone durations) improved statis-tically significantly. A pre/post-test task of phonemic reception, which assessed the identification of phonic elements within words, improved in the case of six of the seven participants (av-erage gain 1.5 years in language development). Their second periment of the same study replicated the results of the first ex-periment, and showed additionally that the performance of a control group playing nonadaptive videogames for the same time period did not improve.

In their second study, language comprehension was similarly trained and ameliorated with acoustically modified speech (Tal-lal et al., 1996) with the same participants as in the first study (Merzenich et al., 1996). The training (3 hours per day, 5 days per week in a laboratory, and additionally 1-2 hours per day, 7 days per week at home; over the time period of 4 weeks) consisted of

stimuli that were temporally prolonged with fast transitions amplified but preserving the spectral content. The stimuli were presented in lengthening, grammatically more complex, sequenc-es. The pre/post-test consisted of standardized tests of speech/

language with unprocessed speech and auditory temporal acui-ty [partly overlapping with the pre/post-tests of (Merzenich et al., 1996)], and indicated an approximate 2 year developmental gain. The second experiment was carried out as in their first study and included a matched control group who trained temporally nonadaptive computer games. A significantly greater improve-ment was shown in the group that received acoustically modi-fied training, and the gains were observed in a retest after six weeks. The difficulty in interpreting the results of these studies (Merzenich et al., 1996; Tallal et al., 1996) is in the overlap be-tween different training methods for the same participants dur-ing the same period of time, in the close correspondence between the task types in the training and pre/post-test studies (e.g. in training to follow complex commands and pre/post Token test), and in the supposedly identical tasks in the pre/post-tests.

4.5.2.2 Developmental dyslexia

Habib and his colleagues (Habib et al., 1999) devised acoustical-ly modified training for (phonologicalacoustical-ly) dyslexic French chil-dren at the ages of 10-12 years (45 min per day for 5 weeks). Half of the 12 children were trained with prolonged speech stimuli with short transients amplified, the other half with unmodified speech. The acoustically modified training led to improvement, which was still observable in post-training after one month.

Again, it is difficult to interpret the results because of the partly identical tasks in pre-training, pre-testing and post-testing; the inherent memory load of the tasks and the presumably resulting gradually improving strategy formation (lists of 3-4 nonwords);

and the possibly intervening effects of the greater attentional/

cognitive demands in the acoustically modified condition.

Temple and her colleagues (Temple et al., 2000) studied three dyslexic adult readers and showed that training (100 min per

day, 5 days per week, for approximately one month) with adap-tive tasks of the FastForword method of Tallal resulted in im-proved speech/nonspeech processing in the case of two partici-pants. In fMRI responses, the behavioral changes were paralleled by increasing left prefrontal activation for rapid transients, which was absent in the dyslexic group before training but observed in the controls. The number of participants was very small and un-fortunately no controls participated in the training..

Recently Kujala, Karma, and their colleagues (Kujala et al., 2001) have assessed the effect of nonlinguistic audiovisual train-ing in dyslexic Finnish 7-year-old children (14 10-min sessions over the time period of 7 weeks). The training consisted of match-ing visually and auditorily (pitch, duration, SOA, intensity var-ied) presented rhythmic patterns. Differences between the groups were observed after the training: the training group read faster and more accurately than the group that did not receive train-ing. Differences in MMN responses to tone order discrimination were observed as between-group differences in the post-training amplitudes, and when the pre/post-training sessions were com-pared, there were within-group differences in response ampli-tude. The pre/post-test changes in reading performance correlat-ed with the changes in the MMN responses. Interpretation is somewhat complicated again by the lack of a control group who would have practiced some nontemporal task to an equivalent extent. The writers themselves also indicate that the MMN am-plitudes in which the changes were observed can be modulated by attention. Attentional factors can easily be affected and re-fined by such training procedures.

Taken together, it seems that both reading and linguistic processing can be somewhat ameliorated with only a short re-hearsal period. However, despite the intriguing findings, it is not fully understood whether or not something really improves with these temporal acuity-training methods. Further, if real improve-ment occurs, it is not completely understood what exactly is im-proved. Is it the underlying relevant neural substrate, or the trained/more general temporal acuity, or some aspect of language processing/reading, or the specific attentional/memory/other

cog-nitive demands in the trained/tested tasks, or performance and strategies in psychometric tasks more generally?