Field strength in neuro-MR imaging: a comparison of 0.5 T and 1.5 T

A study was undertaken comparing neurological magnetic resonance imaging at high (1.5 T) and mid (0.5 T) field strengths. Twenty-eight patients (20 head and 8 spine) from our routine case load volunteered to undergo two consecutive and identical MR studies on the two systems. The two MR systems were built by the same manufacturer and were equipped with essentially identical hardware and software. Individual patient studies were performed consecutively in adjacent MR suites, and pulse sequence parameters were replicated exactly at the two field strengths. One exception to this rule was that the second echo of the long TR sequence in the head was acquired with a narrow receiver bandwidth on the 0.5 T system. The resulting axial double echo long repetition time (TR) and sagittal short TR head images and sagittal short and double echo long TR spine images were graded by two blinded observers (senior staff neuroradiologists) on two levels. First, the images were graded for image quality, i.e., conspicuousness of artifacts and clarity in depiction of normal and pathologic anatomy. Second, diagnostic accuracy of MR was assessed relative to the clinical-pathologic diagnosis in each case. The image quality of the 1.5 T system was rated superior in both the head and spine for most specific items assessed. This observer preference for 1.5 T images did not, however, translate into greater diagnostic accuracy for the 1.5 T system in the head. Although the 1.5 T system did have a slight advantage in diagnostic accuracy in the spine, a significant difference was not found.