Apparent diffusion coefficient measurements in the hippocampus and amygdala of patients with temporal lobe seizures and in healthy volunteers.

PURPOSE

The goals of this work were to measure the apparent diffusion coefficients (ADCs) for both hippocampus and amygdala of persons diagnosed with temporal lobe epilepsy (TLE) and unilateral hippocampus pathology on magnetic resonance imaging and to evaluate the sensitivity of diffusion-weighted (DW) images in determination of the lateralization of the epileptogenic focus.

METHODS

Thirteen cases with a TLE diagnosis and 21 healthy subjects were evaluated. Fluid-attenuated inversion recovery and T2W images of TLE cases revealed hippocampal volume loss and signal intensity changes. DW images were obtained by spin-echo echo-planar sequences vertical to the hippocampal axis. Qualitative and quantitative ADCs for left and right hippocampus and the amygdala of the controls and the patients were determined. Hippocampal ADCs were obtained independently at the head, body, and tail levels of the hippocampus. Statistical evaluation was conducted with Kruskal-Wallis and Mann-Whitney U tests. Predictive cutoff levels of hippocampal ADCs for identifying pathologic areas were established through receiver operating characteristic (ROC) curve analysis.

RESULT

On conventional images, 5 of 13 cases had right hippocampal pathology, and 8 of 13 cases had left hippocampal pathology. There were no bilateral hippocampal changes in signal intensity and no cases with bilateral atrophy. The amygdala was normal in all patients except one case of hyperintense signals. No statistical differences were found between the hippocampal and amygdaloid ADCs of the control subjects (P > 0.05). However, there was a significant difference between the ADCs for the side with hippocampal pathology and the ADCs for the contralateral side, and the control group (P < 0.001). No statistical difference was detected for the amygdala (P > 0.05). Hippocampal and amygdaloid ADCs of the contralateral lesion and the values of the control group were not statistically significantly different (P > 0.05). ROC curve analysis indicated 136 as the best cutoff level for hippocampal pathology.

CONCLUSION

DW trace images are insensitive in lateralization of hippocampal pathology; however, lateralization can be achieved through ADC measurements of the hippocampus. An increase in ADC on the affected side should be considered as indicating pathology. On the other hand, amygdaloid ADC values remain inaccurate.