Frontiers in Psychology (Apr 2015)
What Picture Descriptions Can Reveal about Disordered Communication and the Brain <br />
Abstract
Analysis of descriptions of the “Cookie Theft” picture from the Boston Diagnostic Aphasia Examination have been shown to (1) distinguish between controls, and chronic mild, moderate, and severe aphasia(Craig et al., 1993; Yorkston & Beukelman, 1980); and (2) distinguish distinct profiles of deficits in chronic right hemisphere (RH) stroke (Myers, 1978; Trupe & Hillis, 1985). We hypothesized that analysis of the “Cookie Theft” picture descriptions in acute stroke would also: (1) provide quantitative measures of severity of communication impairment that correlate with volume of infarct or volume of lesion in key regions of interest; and (2) provide quantitative measures of hemispatial attention that distinguish RH from left hemisphere (LH) stroke and from controls. Methods: We analyzed picture descriptions of patients with acute ischemic LH (n=28), RH (n=25) stroke, and healthy controls (n=25) for total content units (CU) produced by previously studied healthy controls during picture descriptions (Yorkston & Beukelman, 1980), syllables/CU, and ratio of left page:right page CU (LCU:RCU). CUs defined as nouns, adjective, and verbs relevant to the picture. Diffusion-weighted images were registered to a common atlas, to measure volume of infarct and percent damage to regions of interest without knowledge of behavioral analysis. We evaluated differences across groups with ANOVA and across pairs of groups with t-tests. We evaluated correlations between each measure and volume of infarct and percent damage to 6 cortical and 3 white matter regions of interest in each hemisphere (inferior frontal gyrus, superior temporal gyrus, middle temporal gyrus, fusiform gyrus, angular gyrus, and supramarginal gyrus; superior fronto-occipital fasciculus, superior longitudinal fasciculus, and sagittal stratum) with Pearson correlations. Results: There were no differences between groups in age, education, or sex distribution. LH and RH patients produced fewer CU (p<0.0001) and more syllables/CU (p=0.026) than controls, and did not differ from each other. RH patients showed lower mean LCU:RCU than LH patients (0.80 vs 1.4; p=0.007), and lower than controls (0.80 vs 1.1; p<0.02). For RH stroke patients, total CU negatively correlated with total lesion volume (Pearson r=-0.51; p<0.003) and percent damage to middle temporal gyrus (r=0.53; p=0.02) and superior fronto-occipital fasciculus (r=-0.66; p=0.002). Syllables/CU correlated with lesion volume (r=0.40; p=0.02) and percent damage to inferior frontal gyrus (r=0.94; p<0.0001) and sagittal stratum (r=0.95; p<0.0001). For LH stroke patients, total CU correlated with percent damage to superior longitudinal fasciculus (r= -0.45; p=0.01) and angular gyrus (r= -0.45; p=0.02). Left:right CU correlated with percent damage to left fusiform gyrus (r=0.58; p=0.001). Conclusions: Patients with LH and RH stroke produce fewer appropriate CU in describing the picture, and are less efficient (produce more syllables/CU) compared to controls. On average, patients with RH stroke mention more CU on the right, and LH stroke patients mention more CU on the left of the page, and this bias correlates with percent damage to areas previously associated with right hemispatial neglect (Kleinman et al., 2007). This test typically takes less than one minute to administer and requires little time for analysis, but provides a rich assessment of right and left cognitive function.
Keywords
