Texture Feature Analysis for Classification of Early-Stage Prostate Cancer in mpMRI
This work addresses the problem of explainability in AI for medical imaging, specifically for radiologists and clinicians, but it is incremental as it focuses on feature analysis rather than introducing new methods.
The study tackled the challenge of interpreting machine learning models for early-stage prostate cancer classification in MRI by analyzing feature contributions, finding that most commonly used features are strongly correlated and have negligible impact, while identifying a small set of key features that determine classification outcomes.
Magnetic resonance imaging (MRI) has become a crucial tool in the diagnosis and staging of prostate cancer, owing to its superior tissue contrast. However, it also creates large volumes of data that must be assessed by trained experts, a time-consuming and laborious task. This has prompted the development of machine learning tools for the automation of Prostate cancer (PCa) risk classification based on multiple MRI modalities (T2W, ADC, and high-b-value DWI). Understanding and interpreting the predictions made by the models, however, remains a challenge. We analyze Random Forests (RF) and Support Vector Machines (SVM), for two complementary datasets, the public Prostate-X dataset, and an in-house, mostly early-stage PCa dataset to elucidate the contributions made by first-order statistical features, Haralick texture features, and local binary patterns to the classification. Using correlation analysis and Shapley impact scores, we find that many of the features typically used are strongly correlated, and that the majority of features have negligible impact on the classification. We identify a small set of features that determine the classification outcome, which may aid the development of explainable AI approaches.