ИСТИНА

Introduction: We analyzed the relation between the sizes of several brain structures (hippocampus, caudate nucleus, cingulate cortex) and different verbal working memory characteristics. Methods: Subjects: 43 right-handed females (mean age – 60.4 years old). We performed magnetic resonance morphometric analysis of 3-dimensional T1 images in order to measure hippocampus and caudate nucleus volumes (in mm3) in both hemispheres. We also calculated absolute square surfaces (in mm2) of three cingulate cortex regions – anterior (Brodmann areas 24 and 33), posterior ventral (Brodmann area 23), posterior dorsal (Brodmann area 31) areas. Verbal working memory characteristics (memory capacity, permanency and the amount of different memory errors) were assessed by Luria's neuropsychological tests modified by J.M. Glozman (1999). We calculated non-parametric correlations (Spearman coefficient) between individual verbal memory characteristics and morphometrical data. Statistically significant (p < 0,05) correlations were analyzed further. Results: We revealed correlations between verbal information maintenance permanency and the left caudate nucleus volume (r = 0.39); associative memory capacity correlates positively with the left hippocampus volume (r = 0.31). Decline in the amount of verbal memory errors (confabulations) is related to the increased size of the left anterior cingulate cortex (r = -0.35). Furthermore, verbal working memory capacity (r = -0.35) and permanency (r = -0.33) correlate negatively with the size of the left posterior dorsal cingulate cortex, whereas the amount of fluctuations (r = 0.46) and word replacements (r = 0.32) correlate positively with the increased size of this brain region. Conclusions: According to the obtained data, the volume of caudate nucleus in left hemisphere correlates with the permanency of verbal information maintenance after the interfering stimuli in different memory tests but not with verbal memory capacity. Thus, we can suggest that the role of this brain structure in verbal memory is the resistance to external interference. Left anterior cingulate cortex is related to resistance to confabulations; the increase of this brain area leads to the confabulations decline in the semantic coding test. We suggest that left anterior cingulate cortex in verbal memory functioning suppresses irrelevant information and resists the internal semantic interference. Left posterior dorsal cingulate cortex size correlates negatively with verbal working memory capacity and permanency. It also correlates positively with the amount of different memory errors. We suggest that the role of this brain region consists of relevant information subtraction. Moreover, left hippocampus volume correlates positively with associative memory capacity. It may point to hippocampus involvement in information consolidation by the creation of associative links between memory traces. We can hypothesize that verbal information maintenance in the working memory is provided by the functioning of five sub-systems: 1. Sensory filter (partial filtration of the extra sensory information at the moment of the information maintenance in the memory. This block functioning is probably related to the left caudate nucleus activity, which slows down the thalamic nuclei (responsible for the entrance of sensory information into the left hemisphere of the brain)); 2. Sensory screen (temporary verbal information storage); 3. Semantic filter (the suppression of internal interference. Its functioning is probably related to the left anterior cingulate cortex activity); 4. Central executive (voluntary control of incoming information by the regulation of filters functioning); 5. Archiver (information collection and classification for the long-term storage by associative links creation. Supposedly, this sub-system functions because of the left posterior cingulate cortex and left hippocampus activity).