Approximately a third of stroke patients have aphasia.As the most common clinical manifestation of aphasia patients, speech communication difficulties will lead to emotional disorders, reduce communication participation and even reduce quality of life. Distinct from dementia patients, aphasia patients after stroke are still able to discriminate and recognize the acquired semantic knowledge but cannot extract randomly and name correctly compared with healthy people. It has been proved that semantic treatment can maintain a longer duration of efficacy than phonological treatment in speech therapy. Semantic feature analysis therapy has been widely used in clinical practice and achieved good behavioral therapeutic effects. However it remains unknown whether the impaired brain function will be restored. To explore the semantic processing of human brain, it is necessary to identify " what" category of semantic features have been activated, " where" the activation center and related activities can be seen, " when" generates activities and affect " how" processing by describing the time sequence. In order to explore the neuroimaging therapeutic effects of training materials with specific semantic features, the paper reviews the influence of semantic features in spatial, temporal and frequency relationships on the brain functional activity, and prospect the future research and clinical application.

Objective:To explore the temporal and spatial changes of the whole brain during the picture naming tasks and the influence of different semantic similarity in the blocked-cyclic paradigm.Methods:Ten healthy right-handed adults were selected.The blocked-cyclic paradigm is divided into two naming tasks consist of related and unrelated pictures.The brain activation of the whole brain in different time windows was recorded by magnetoencephalography(MEG) during the picture naming tasks.Results:In the visual-related time windows, the dominant regions appeared in the related picture naming task, located in left frontal pole (1.421±0.468), right frontal pole (1.431±0.435), left orbitofrontal gyrus(1.601±0.620), right orbitofrontal gyrus (1.567±0.556), left superior frontal gyrus (1.899±0.062), right superior parietal gyrus (5.218±0.678) and left lingual gyrus (5.016±0.088) compared with the unrelated picture naming task (1.038±0.217, 1.131±0.235, 1.312±0.316, 1.253±0.340, 1.710±0.151, 4.538±0.478, 4.275±0.251), the differences were statistically significant ( P<0.05). In the semantic-related time windows, the dominant regions appeared in the unrelated picture naming task, located in left paracentral lobule (3.176±0.392), right cuneus (4.190±0.330), left posterior cingulate (2.430±0.196) and right posterior cingulate (2.405±0.236) compared with the related picture naming task (2.594±0.254, 3.626±0.560, 2.038±0.115, 1.990±0.094), and the differences were statistically significant ( P<0.05). In the phonological-related time windows, the dominant regions appeared in the unrelated picture naming task, located in left superior temporal sulcus (3.709±0.274), left superior temporal gyrus (3.901±0.380), left middle temporal gyrus (3.340±0.380), left transverse temporal gyrus (4.449±0.265), left supramarginal gyrus (3.205±0.308), left insula (3.140±0.204), left anterior cingulate (2.217±0.194) and right anterior cingulate (2.341±0.167) compared with the related picture naming task (2.662±0.300, 2.503±0.342, 2.614±0.324, 2.633±0.281, 2.663±0.278, 2.248±0.284, 1.818±0.315, 2.005±0.437), and the differences were statistically significant ( P<0.05). Conclusion:In different time windows, semantic similarities can lead to differences in brain activation.The related picture naming task can cause the dominant activation of bilateral prefrontal lobe in the visual processing stage, which may be related to sustained attention and category judgment while the unrelated picture naming task can cause dominant activation of the left temporal gyrus, left insula and bilateral limbic system during speech processing, which corresponds to semantic coding and phonology introspection.