Activity in preserved left hemisphere regions predicts anomia severity in aphasia

Abstract

Understanding the neural mechanism that supports preserved language processing in aphasia has implications for both basic and applied science. This study examined brain activation associated with correct picture naming in 15 patients with aphasia. We contrasted each patient's activation to the activation observed in a neurologically healthy control group, allowing us to identify regions with unusual activity patterns. The results revealed that increased activation in preserved left hemisphere areas is associated with better naming performance in aphasia. This relationship was linear in nature; progressively less cortical activation was associated with greater severity of anomia. These findings are consistent with others who suggests that residual language function following stroke relies on preserved cortical areas in the left hemisphere.

Figure 1.

Brain activation associated with naming pictures in the neurologically healthy control group. The color scale represents Z-scores compared with baseline (viewing abstract pictures). The left hemisphere is shown on the left.

Figure 2.

The number of total and correct naming attempts by each brain-injured participant. The height of each column denotes the number of naming attempts, whereas correct naming is shown in a darker shade of gray.

Figure 3.

Cortical areas associated with naming task performance (red–yellow scale) in patients with aphasia. The lesion overlay map for all 15 patients is shown in the blue–green scale (note that the maximum range of the scale is set at 10, reflecting the highest degree of lesion overlap). The slight overlap between the activation and lesion maps (at the anterior–inferior and posterior edges of the lesion map) is shown in bright pink.

Figure 4.

The relationship between intensity of activation (x-axis; measured in Z-scores compared with a group of normal control participants) and the number of correct naming attempts (y-axis; out of 80 pictures) during fMRI scanning. The dotted line represents a line of best fit.

Figure 5.

Critical brain damage associated with lower brain activation in the VOI generated in the third-level analysis. This analysis identifies regions where injury predicts lower Z-scores (as shown in the horizontal axis of Fig. 4). This analysis used overall lesion volume as a nuisance regressor. The resulting statistical map is shown uncorrected for multiple comparisons. Therefore, injury to Broca's area (shown here in green) predicted reduced brain activation in the brain areas shown as orange in Figure 3.