Quality Assurance Failure in MRI Simulation

Possible causes: 1) SNR decrease: coil failure or degradation (e.g., preamplifier, cable), increased noise from the environment (e.g., nearby equipment), detuning of the coil, or a change in the receive bandwidth. 2) Uniformity variation: coil element failure, shimming issues, or a problem with the RF transmit system. 3) Geometric distortion: gradient nonlinearity, main field inhomogeneity, or a problem with the gradient coils (e.g., eddy currents). Also, the phantom itself may have degraded (e.g., fluid evaporation, air bubbles). To investigate: 1) Review the QA records to see when the changes occurred. 2) Perform a more detailed coil test (e.g., coil element check with a phantom). 3) Check the scanner's logs for errors. 4) Test with a different coil if available. 5) Measure the main magnetic field homogeneity with a shim scan. 6) Perform a gradient calibration check. 7) Inspect the phantom for damage or fluid loss. 8) Compare with the baseline images to see the pattern of distortion. 9) Contact the vendor for service. 10) Consider environmental changes (e.g., new equipment nearby).

Acceptable limits for MRI simulation are not standardized but are based on AAPM Report No. 100 and institutional protocols. Typically, SNR should be within 10-20% of baseline; uniformity should be within 10-15%; geometric distortion should be less than 2 mm for treatment planning, but for stereotactic cases, less than 1 mm. In this case, SNR decreased by 20% (borderline), uniformity variation 15% (borderline), and geometric distortion 2 mm at periphery (concerning for large fields). To determine acceptability: 1) Compare with the baseline values and tolerances set during acceptance. 2) Assess the impact on clinical images: review recent patient scans for artifacts or distortion. 3) For geometric distortion, perform a more detailed mapping with a large field-of-view phantom and compare with CT. 4) If distortion is only at the periphery and the clinical target is central, it may be acceptable, but for head and neck, the periphery may include critical structures. Actions: 1) If within tolerance, continue use but monitor closely. 2) If outside, take the scanner out of clinical service for simulation until repaired. 3) Schedule vendor service. 4) Notify the radiation oncologists and physics team. 5) Use CT simulation in the interim. 6) Document the issue and corrective actions. 7) Re-baseline after repair.

To better detect issues: 1) Increase the frequency of geometric distortion testing to monthly, using a dedicated phantom (e.g., large FOV phantom) and analyze distortion across the entire FOV. 2) Add a coil element check to monthly QA (e.g., using a coil phantom or checking each element's signal). 3) Include a test for main field homogeneity (e.g., phase map) monthly. 4) Perform a more comprehensive SNR test with multiple coils and sequences. 5) Add a test for gradient calibration (e.g., using a diffusion phantom). 6) Implement a trend analysis of QA metrics to detect gradual changes. 7) Use a phantom that simulates human anatomy for more clinically relevant testing. 8) Consider annual testing with a large FOV phantom for distortion. 9) Document and review any changes in scanner hardware or software. 10) Train technologists to report any perceived image quality changes.