Impact of Magnetic Field Strength on Radiomics-Based Knee Injury Diagnosis

Motivation: MRNet is a widely used database for knee injury diagnosis. While several classifiers achieve high accuracy on internal datasets, their performance often declines on external data. Researchers primarily attributed this degradation to site effects, though the role of magnetic field strength (MFS) variations has likely been underestimated. Goals: To disentangle the effects of MFS in knee injury classification. Approach: MFS was inferred using a Gaussian mixture model. Radiomic features were evaluated through one-way ANOVA to quantify MFS effects. Results: Preliminary results suggest that MFS substantially influences radiomic feature distributions and introduces bias, particularly in imbalanced datasets.