Increasing patient throughput in magnetic resonance imaging: A practical approach

A major factor governing the throughput of patients in a magnetic resonance imaging (MRI) service is the number and length of sequences employed. This study investigated the feasibility of prospectively selecting patients for whom a pre-planned examination of the brain, spine or knee could be applied with a high chance of success. The implications on throughput of using these pre-planned examinations were determined. 173 patients were studied. A successful outcome was defined as an examination in which a diagnosis could be made with 100% confidence and without the need for further sequences. Examinations of the brain (n=113), knee (n=23) and lumbar spine (for disc degeneration, n=14) were performed with success rates of 96%, 87% and 64%, respectively. Examinations of the lumbar spine (for radicular symptoms, n=20) and cervical spine (n=3) were performed with success rates of 85% and 66%, respectively. The examinations of the brain, knee and lumbar spine (for disc degeneration) were completed within the target time of 20 min in 92%, 95% and 69% of cases, respectively. Examinations of the lumbar spine (for radicular symptoms) and cervical spine were completed within 30 min in 75% and 33% of cases, respectively. Analysis of our results suggests that success rates could be improved by minor changes in sequence design. Carefully selected patients can be successfully examined in short examination times. Instituting these practices would increase the efficiency of MR machine time and improve patient throughput by 80-125%. This would have a marked effect on the length of waiting lists, and would increase the general availability of MRI.