Analysis and evaluation of selected magnetic resonance imaging variability factors using a cervical spine reference phantom

Magnetic Resonance Imaging (MRI) is the clinical imaging modality of choice in many diagnostic imaging studies, including the spine. Diagnostic cervical spine MRI image quality depends on both scanner and patient variability factors. The source of poor image quality is important information and very useful during clinical evaluation.; After conducting a retrospective review of 25 neuroradiologic MR cervical spine patient examination case studies, a cervical spine reference phantom was designed and fabricated at the Medical College of Ohio for use with a clinical GE Signa 1.5 Tesla whole-body MRI scanner and a GE Cervical/Thoracic/Lumbar (CTL) array surface coil. The reference phantom was tested and its performance validated prior to data collection. Approximately 30 patients were imaged using the reference phantom during routine clinical neuroradiologic examination. Approximately 158 cervical spine images were collected, evaluated, and analyzed. The performance of the reference phantom was evaluated by a team of radiologists and multiple software tools.; The primary result of this research was the creation of a film-based MRI reference phantom that was designed and tested confirming its performance and operation. The research correlated subjective results with the reference phantom measurements and performance. The following five parameters were evaluated: (1) spatial resolution, (2) contrast, (3) position, (4) neck size, and (5) signal-to-noise ratio (SNR). A significant sample of degraded spin-echo and gradient-echo MR images were collected and analyzed. Using both quantitative and qualitative techniques these images were evaluated. Descriptive tests and statistical analyses were performed on all factors evaluated. A Student's t-test was performed to test the differences between the images that were assessed to be above average quality and those that were assessed to be below average quality. A total of 14 separate t-tests were completed. Patient-induced variability factors were found to be the primary cause of perceived image quality degradation. Scanner-induced variability factors did not contribute significantly to the image quality degradation observed. The results and findings of this research identified areas where image quality may be improved and where further research might be directed.