Multiparametric radiomics methods for breast cancer tissue characterization using radiological imaging

Area of Science:

• Radiology and Medical Imaging
• Quantitative Imaging
• Biomarkers

Background:

• Multiparametric radiological imaging is crucial for disease detection and diagnosis.
• Current radiomics methods are limited to single imaging parameters, restricting information capture.
• This limitation hinders the integration of radiomics into clinical settings.

Purpose of the Study:

• To develop a multiparametric radiomic (mpRad) framework.
• To extract first and second-order radiomic features from multiparametric datasets.
• To overcome the limitations of single-image radiomics.

Main Methods:

• Developed five radiomic techniques analyzing inter-voxel and inter-parametric relationships.
• Utilized multiparametric magnetic resonance imaging (MRI) breast datasets from 138 patients (97 malignant, 41 benign).
• Assessed diagnostic performance using sensitivity, specificity, ROC, and AUC analysis.

Main Results:

• The mpRad framework achieved 82.5% sensitivity and 80.5% specificity in classifying malignant vs. benign breast lesions.
• Achieved an Area Under the Curve (AUC) of 0.87 (0.81-0.93).
• mpRad demonstrated a 9–28% increase in AUC compared to single radiomic parameters.

Conclusions:

• Introduced the novel mpRad framework for enhanced radiomic analysis.
• Extended radiomics from single images to multiparametric datasets.
• Improved characterization of underlying tissue biology for better clinical application.