A new method for evaluation of mild traumatic brain injury with neuropsychological impairment using statistical imaging analysis for Tc-ECD SPECT.

OBJECTIVE

The objective of this study was to identify specific brain lesions with regional perfusion abnormalities possibly associated with neuropsychological impairments (NPI), as sequela after mild traumatic brain injury (MTBI), using 99mTc-ethylcysteinate dimer single photon emission computed tomography (Tc-99m ECD SPECT) and its novel analytic software.

METHODS

We studied 23 patients with diffuse axonal injury with NPI group (Impaired-DAI), 26 with MTBI with NPI group (Impaired-MTBI) and 24 with MTBI without NPI group (Healthy-MTBI). In each subject, Tc-99m ECD SPECT images were analyzed by easy Z score imaging system (eZIS) and voxel-based stereotactic extraction estimation (vbSEE). Segmented into lobule levels, ROIs were set in 140 areas in whole brain, and relative regional low Tc-99m ECD uptake was computed as "extent" (rate of coordinates with Z score >2.0 in the ROI). Receiver operating characteristic analysis was performed using "extent" to discriminate the three groups.

RESULTS

The highest area under the curve (AUC) value for data of Impaired-DAI and Healthy-MTBI groups was obtained in ROI on the left anterior cingulate gyrus (LtACG), with AUC of 0.93, optimal "extent" cutoff value of 10.9%, sensitivity 87.0%, specificity 83.3%. The highest AUC value for data of Impaired-MTBI and Healthy-MTBI groups was also in the LtACG, with AUC of 0.87, optimal "extent" cutoff value of 9.2%, sensitivity 73.1%, specificity 83.3%.

CONCLUSIONS

Using two analytic software packages, eZIS and vbSEE, we identified specific lesions with low regional Tc-99m ECD uptake possibly associated with NPIs after MTBI. Especially, this trend was most marked in the left anterior cingulate gyrus in MTBI patients with NPIs and those with DAI. The optimal "extent" cutoff value, as a criterion for SPECT abnormality, might help the diagnosis of NPIs after MTBI.