The Use of Quantitative Metrics and Machine Learning to Predict Radiologist Interpretations of Image Quality and Artifacts

A dataset of 3D-GRE and 3D-TSE brain 3T post contrast T1-weighted images as part of a quality improvement project were collected and shown to five neuro-radiologists who evaluated each sequence for image quality and artifacts. The same scans were processed using the MRQy tool for objective, quantitative image quality metrics. Using the combined radiologist and quantitative metrics dataset, a decision tree classifier with a bagging ensemble approach was trained to predict radiologist assessment using the quantitative metrics. The resulting AUCs for each classification task were above 0.7 for all combinations of sufficiently represented classes and qualitative image metrics.