A phantom study to assess the accuracy of stereotactic localization, using T1-weighted magnetic resonance imaging with the Leksell stereotactic system.

This phantom study assesses the accuracy of stereotactic localization using the Leksell G frame (Elekta Instruments AB, Stockholm, Sweden) with T1-weighted magnetic resonance imaging (Siemens 1.5 T Magnetom; Erlangen, Germany). The coordinates of an array of solid perspex rods were determined and compared with measured values in a series of transverse, coronal, and sagittal images. The maximum absolute errors observed (X = 2.7 mm, Y = 7.0 mm, Z = 8.0 mm) were discouraging. However, the more reasonable mean errors (X = 0.4 mm, Y = 0.7 mm, Z = 1.3 mm) reflect considerable variation in accuracy throughout the volume assessed and limitation of maximum errors to specific areas. We present details of the spatial variation and discuss possible mechanisms for improving accuracy. The overall results are of direct relevance only to the scanner used. These results are, however, an indication of the need to approach with caution stereotactic localization using magnetic resonance imaging and to emphasize the requirement for quality assurance and for a comprehensive study of the scanner's characteristics.