Patient-specific radiomic feature selection with reconstructed healthy persona of knee MR images

TL;DR

This work tackles the interpretability gap in knee MRI pathology classification by integrating radiomic features with patient-specific feature weighting and a healthy persona generated by a denoising diffusion model. The method learns to weight and select radiomic features for each patient, then uses logistic regression for classification, while augmenting the feature pool with a patient-specific healthy baseline to enable more robust feature discovery. Empirical results on MRNet knee MRI data show competitive or superior performance to state-of-the-art deep learning approaches across general abnormalities, ACL tears, and meniscus tears, with added interpretability through explicit feature selection and pathology-free baselines. The framework advances clinically actionable radiomic analysis by combining subject-specific feature weighting, generative healthy personas, and interpretable downstream classification, with code publicly available for reproducibility.

Abstract

Classical radiomic features have been designed to describe image appearance and intensity patterns. These features are directly interpretable and readily understood by radiologists. Compared with end-to-end deep learning (DL) models, lower dimensional parametric models that use such radiomic features offer enhanced interpretability but lower comparative performance in clinical tasks. In this study, we propose an approach where a standard logistic regression model performance is substantially improved by learning to select radiomic features for individual patients, from a pool of candidate features. This approach has potentials to maintain the interpretability of such approaches while offering comparable performance to DL. We also propose to expand the feature pool by generating a patient-specific healthy persona via mask-inpainting using a denoising diffusion model trained on healthy subjects. Such a pathology-free baseline feature set allows further opportunity in novel feature discovery and improved condition classification. We demonstrate our method on multiple clinical tasks of classifying general abnormalities, anterior cruciate ligament tears, and meniscus tears. Experimental results demonstrate that our approach achieved comparable or even superior performance than state-of-the-art DL approaches while offering added interpretability by using radiomic features extracted from images and supplemented by generating healthy personas. Example clinical cases are discussed in-depth to demonstrate the intepretability-enabled utilities such as human-explainable feature discovery and patient-specific location/view selection. These findings highlight the potentials of the combination of subject-specific feature selection with generative models in augmenting radiomic analysis for more interpretable decision-making. The codes are available at: https://github.com/YaxiiC/RadiomicsPersona.git

Figures (2)

• Figure 1: Overview of the radiomic feature selection framework (details in Sec.\ref{['sec:method']}).
• Figure 2: Representative cases of ACL and Meniscus Tears (details in Sec.\ref{['sec:results']}).