MAPS: A Quantitative Radiomics Approach for Prostate Cancer Detection

Area of Science:

• Medical Imaging
• Radiomics
• Oncology

Background:

• Prostate cancer diagnosis relies heavily on multiparametric MRI (mpMRI).
• Current automated classification methods may lack interpretability for clinicians.
• Radiomics offers a quantitative approach to extract features from medical images.

Purpose of the Study:

• To develop and evaluate a quantitative radiomics feature model for prostate cancer detection using mpMRI.
• To enhance diagnostic interpretability by grouping radiomics features into clinically relevant categories (MAPS).
• To compare the performance of the proposed radiomics model against conventional methods.

Main Methods:

• A novel tumor candidate identification algorithm was developed.
• A comprehensive radiomics feature model was constructed, grouping features into Morphology, Asymmetry, Physiology, and Size (MAPS) categories, inspired by PI-RADS guidelines.
• Classifiers were trained using clinical mpMRI data from 13 men with confirmed prostate cancer.

Main Results:

• The proposed radiomics-driven feature model demonstrated superior classification performance compared to individual feature groups.
• The model outperformed a comparable conventional mpMRI feature model.
• Preliminary results suggest enhanced diagnostic accuracy and interpretability.

Conclusions:

• The developed radiomics feature model provides an interpretable and effective tool for prostate cancer detection using mpMRI.
• The MAPS feature grouping aligns with established radiological diagnostic criteria.
• Further validation with larger datasets is warranted to confirm findings.