Results of a prospective multicenter study for evaluation of the diagnostic quality of an open whole-body low-filed MRI unit. A comparison with high-field MRI measured by the applicable gold standard

Objective: To evaluate the diagnostic quality of an open whole-body low- field MRI scanner compared to high-field scanners. Materials and methods: Over a period of 3 months, 401 patients with diseases of the kidney (n=78), the shoulder (n=122), the spine (n=105) and the cerebrum (n=96) were prospectively evaluated in four participating centers. They all underwent clinical evaluation, low-field and high-field MRI examination and surgical or follow-up confirmation of diagnosis. Clinical, histopathologic, high-field and low-field MRI diagnoses were recorded in standardized questionnaires that were centrally evaluated. Statistical evaluation comprised two parts: ROC analysis asseased accuracy of MRI and clinical diagnoses; furthermore rates of concordance of high- and low-field MRI diagnosis were calculated. Results: We found no statistically relevant difference in high-field MRI diagnosis compared to low-field MRI diagnosis accuracy measured by clinical or surgical gold standard in three of the four regions examined; in cerebral examinations there was a small yet significant advantages for the high-field systems (P=0.01). Conclusion: We conclude that the open low-field scanner we evaluated using clinical and surgical gold standard as reference is able to achieve comparable diagnostic accuracy compared to high-field scanners at lower costs and greater patient comfort. Limitations due to field strength (signal-to-nose ratio, resolution, scan time) seem to be relevant only in a very small number of cases that warrant high-field examination.