Geometric accuracy in three-dimensional coordinates of Leksell stereotactic skull frame with wide-bore 1.5-T MRI compared with conventional 1.5-T MRI.

Abstract

The use of 1.5-tesla (T) magnetic resonance (MR) imaging with a wide and simultaneously short bore enhances patient comfort compared with traditional 1.5-T MR imaging and is becoming increasingly available in stereotactic radiosurgery treatment planning. However, the geometric accuracy seems unavoidably worse in wide-bore MR imaging than in conventional MR imaging. We assessed the geometric distortion of the stereotactic image attached on a Leksell skull frame in conventional and wide-bore 1.5-T MR imaging. Two kinds of acrylic phantoms were placed on the skull frame and were scanned using computed tomography (CT) and conventional and wide-bore 1.5-T MR imaging. The three-dimensional coordinates on both MR imaging were compared with those on CT. Deviations of measured coordinates at selected points (x = 50, 100, 150 mm; y = 50, 100, 150 mm) were indicated on different axial planes (z = 50, 75, 100, 125, 150 mm). The differences of coordinates were less than 1.0 mm in the entire treatable area for conventional MR imaging. With the large bore system, the differences of the coordinates were less than 1.0 mm around the center but substantially exceeded 1.0 mm in the peripheral regions. Further study is needed to increase the geometric accuracy of wide-bore MR imaging for stereotactic radiosurgery treatment planning.