Language dominance in partial epilepsy patients identified with an fMRI reading task

Area of Science:

• Neuroimaging
• Neurology
• Epilepsy

Background:

• Functional MRI (fMRI) reliably detects frontal language areas but shows inconsistent temporal lobe activation.
• Previous studies have not compared clinical visual judgment with quantitative region of interest (ROI) analysis for temporal language areas.

Purpose of the Study:

• To identify temporal language areas in patients with partial epilepsy using a reading paradigm.
• To compare clinical visual interpretation with quantitative ROI analysis of fMRI data.

Main Methods:

• Thirty temporal lobe epilepsy patients (aged 8-56) underwent 1.5-T fMRI using a silent object naming task versus a visual control.
• Data were analyzed using ROI analysis of t-maps, calculating asymmetry indices (AI) for language dominance (>0.20).
• t-Maps were visually rated by three readers, and results compared with intracarotid amobarbital test (IAT) in 21 patients.

Main Results:

• The fMRI reading task successfully lateralized language in 27 of 30 patients (ROI analysis).
• Twenty-five patients showed left dominance, two right, one bilateral; IAT and fMRI results showed high agreement.
• Interrater reliability for visual and ROI analysis was high (Cramer V 0.77-0.82), with comparable agreement to IAT (Cramer V 0.71-0.77).

Conclusions:

• An fMRI reading paradigm effectively identifies language dominance in both frontal and temporal regions.
• Clinical visual interpretation of fMRI data is a reliable method, comparable to quantitative ROI analysis.