Studies Of Speech Processing In Chinese With Functional MRI

Objective: A fast event-related functional Magnetic Resonance Imaging (fMRI) study was conducted to identify regions that are involved in speech processing in Chinese with a reversed-speech method.Methods: The paradigm followed that of Binder et al. Fourteen subjects were scanned during an auditory task involving stimulus presentation of two characters Chinese words and their reversed format (playing the speech waveform backwards). White noise segments were used as the control condition. Different kinds of stimuli were randomly intermixed and presented in short periods lasting 2 to 10 seconds with a mean of 6 seconds. There was a 50ms blank interval separating each speech segment from the next. Subject briefly pressed a button with their left hand whenever a block of sound segments began or ceased. Whole brain acquisition was conducted on a 1.5 Philips scanner. All analysis followed the standard procedure of SPM2. Activations were reported at a height threshold of p<.0005 (uncorrected), and a cluster threshold of greater than 10 voxels.Results & Discussion: The results show that: processing of noise, reversed words and words activated a highly comparable network of brain regions, including the superior temporal region bilaterally, the left middle temporal region, right fontal region, bilateral supper motor area,cerebellum, right middle cingulum region, postentral -precentral gyms of front lobe bilaterally. Activation related to speech processing was revealed from the reversed words relative to noise comparison, shown in Figure left panel and including superior (BA48) and middle (BA21) temporal region bilaterally, and right pole superior temporal (BA38). Activation related to semantic activation was revealed in the words relative to reversed words contrast shown in Figure right panel and they included the right middle frontal gyrus (BA8/46). No area shows more activation for the reversed condition than for the word condition. Our study replicated findings from Binder et al. that in temporal lobe there is a specialized region responding preferentially to speech sounds when compared with simple nonspeech sounds. Binder and others think the response selectivity is unlikely to be due to activation of lexical-semantic associations of the speech sounds but reflect processing of complex acoustic feature combinations. This point is also confirmed in the present study as we found no activation in temporal lobe but in right middle frontal gyms (BA8/46) related to semantic process.Conclusions: The findings of the present study identify the importance of superior temporal region bilaterally in speech processing. As the first piece of fMRI research on Chinese with the reversed speech methodology, our results support the generality of the speech processing function across languages.