Bringing Objectivity to Cardiac MRI Image Quality: An In-House Audit Across Multi-Vendor Scanners

Background: The evaluation of cardiac magnetic resonance (CMR) images is fundamental to ensuring diagnostic accuracy and optimizing clinical practice. However, image quality (IQ) assessment is often influenced by subjective interpretation, which may introduce variability. To overcome this limitation, an objective IQ audit is required, enabling standardized and reproducible evaluation of image quality. Such an approach ensures consistency across institutions and promotes reliable imaging outcomes, independent of scanner field strength or vendor.

Methods: An in-house objective image quality (IQ) audit was developed by the audit lead for implementation across a multi-site CMR service, incorporating scanners from different vendors and field strengths. For each audit cycle, a predefined number of scans are randomly selected within a set timeframe and evaluated quarterly. To minimize variability, two auditors independently assess each study, and the final score is derived from the median of their evaluations. Image quality is assessed according to SCMR guidelines, applied to standard long-axis and short-axis cine and late gadolinium enhancement sequences, as well as survey and planning images common to all CMR studies. Each projection is evaluated across predefined categories, including correct positioning, presence and mitigation of artefacts, correct coverage for multi-slice sequences, and adequacy of visualization, acknowledging that anatomical variance may affect standard positioning. Points are awarded for each category when criteria are met, with a maximum cumulative score of 100. Additional projections (e.g., RVOT1 and RVOT2 in suspected ARVC cases) are evaluated separately using the same categories but with a percentage-based scoring system. The final median score assigns each study to one of five categories: high standard, good standard, average, poor standard, or very poor standard.

Results: High-quality imaging was consistently achievable across scanners of varying vendors and field strengths. Distribution across the five-point scale demonstrated strong reproducibility of the criteria. Specialized protocol scoring highlighted variability in technically demanding acquisitions, particularly stress perfusion and RVOT views. The audit cycle enabled targeted feedback and quality monitoring, supporting service development during periods of workforce expansion.

Conclusion: The implementation of an objective, standardized CMR image quality audit across a multi-site service proved feasible and reproducible, independent of scanner vendor or field strength. The framework enabled consistent benchmarking of image quality, highlighted areas requiring optimization, and provided a structured basis for targeted training. Routine application of this audit may contribute to improved consistency, diagnostic reliability, and overall quality of CMR practice.

Viability Score Card
Extra Sequences Score Card
Final Results