Improved Detection of Epileptic Foci with PET

Abstract

Objective: This study was conducted to determine the usefulness of a normal reference value database for the detection of areas of abnormal regional cerebral glucose metabolism. The results of this quantitative analysis were compared to a qualitative analysis based only on the detection of left/right asymmetry.

Methods: Ten patients with medically refractory partial complex epilepsy were studied. Metabolic rate images were generated from the patient’s plasma glucose level, FDG clearance in serial arteriolized venous blood samples and PET images calibrated to a well counter.

Results: Using a threshold of greater than 20% left/right asymmetry, we found areas of hypometabolism in temporal regions of four patients. Using a threshold of 1.5 s.d. less than the mean normal value for a particular region, we found areas of hypometabolism in temporal regions of eight patients.

Conclusion: We found that when images were evaluated based on a laboratory normal value database, we were able to detect twice as many areas of hypometabolism.