Assignment Confidence in Localization of the Hand Motor Cortex: Comparison of Structural Imaging With Functional MRI

OBJECTIVE. The purpose of this study was to assign confidence levels to structural MRI and functional MRI (fMRI) for localization of the primary motor cortex. MATERIALS AND METHODS. Ninety-one fMRI studies with at least one motor task (178 hemispheres) were identified. Three anatomic assessments were used to localize the primary motor cortex: relation between the superior frontal sulcus and precentral sulcus; cortical thickness; and configuration of the precentral knob. In 105 hemispheres, interreader agreement was assessed for two investigators with different experience levels. Confidence ratings from 0 to 5 (0, no confidence; 5, 100% confidence) were assigned for fMRI and each anatomic localization method. RESULTS. Cortical thickness had the highest confidence rating (mean, 4.90 +/- 0.47 [SD]) with only one failure. The relation between the superior frontal sulcus and precentral sulcus had the lowest confidence rating (4.33 +/- 0.91) with three failures. The greatest statistical significance was observed for the cortical thickness and superior frontal sulcus-precentral sulcus methods (post hoc Bonferroni test, p < 0.001). Confidence rating scores were significantly higher for the cortical thickness sign than for fMRI results (4.72 +/- 0.54) for a single motor task (post hoc Bonferroni test, p = 0.006); however, the mean confidence rating for fMRI improved to 4.87 +/- 0.36 when additional motor tasks were performed. Interreader differences were least for the cortical thickness sign (paired t test, t = 4.25, p < 0.001). CONCLUSION. Cortical thickness is a better anatomic landmark than fMRI localization for assigning confidence regarding localization of the primary motor cortex; however, localization of motor function is more specific when combined with fMRI findings. Multiple techniques can be used to increase confidence in identifying the hand motor cortex.