Evaluation of a prospective scoring system designed for a multicenter breast MR imaging screening study

Area of Science:

• Radiology and Medical Imaging
• Oncology
• Biomedical Engineering

Background:

• Magnetic resonance imaging (MRI) is crucial for breast cancer screening in high-risk populations.
• Accurate characterization of breast lesions identified on MRI is essential for effective screening.
• Existing classification systems require evaluation for their utility in high-risk screening settings.

Purpose of the Study:

• To prospectively assess the accuracy of a novel lesion classification system for breast MRI screening.
• To evaluate the system's performance in distinguishing malignant, suspicious, and benign breast lesions.
• To identify key features contributing to accurate lesion characterization in high-risk women.

Main Methods:

• Analysis of 1541 contrast-enhanced breast MRI examinations, including 1441 screening exams in high-risk women.
• A five-feature scoring system (morphology, pattern of enhancement, maximal focal enhancement, signal intensity-time ratio, pattern of contrast material washout) was developed and tested.
• Diagnostic accuracy was evaluated using receiver operating characteristic (ROC) curve analysis and interreader agreement (kappa statistics).

Main Results:

• The overall lesion score achieved an area under the ROC curve (AUC) of 0.88.
• Morphology, pattern of enhancement, and pattern of contrast material washout were the most effective components.
• A simplified scoring system demonstrated comparable diagnostic accuracy (sensitivity 84%, specificity 86.0%) to the original system.

Conclusions:

• Simple qualitative descriptors of lesion enhancement, particularly pattern of contrast material washout, provide good diagnostic accuracy.
• Quantitative enhancement parameters (maximal focal enhancement, signal intensity-time ratio) were less useful for lesion characterization in this screening context.
• The developed classification system is accurate and practical for use in high-risk breast cancer screening with MRI.