High scattering sensitivity entropy imaging for breast tumor characterization and classification.

Area of Science:

• Medical imaging
• Biomedical engineering
• Diagnostic radiology

Background:

• Breast lesion diagnosis is challenging, with ultrasound limited by image quality and operator experience.
• Improving ultrasound image analysis is crucial for accurate breast tumor characterization.

Purpose of the Study:

• To introduce a high scattering sensitivity fuzzy entropy (FE) imaging method for enhanced breast tumor detection.
• To improve image contrast and lesion detectability using FE imaging.
• To enable preliminary classification of breast lesions as benign or malignant via quantitative analysis of ultrasound data.

Main Methods:

• A sliding window approach was used to calculate entropy values for each pixel, creating a parametric image.
• Clinical experiments involved classifying lesions by biopsy and calculating average entropy values for benign and malignant tumors.
• Statistical analysis, including one-sample t-tests and Tukey tests, was performed to assess differences in entropy values.

Main Results:

• The FE method achieved a high Matthews correlation coefficient (MCC) of 0.875 and F1 score of 0.876.
• FE imaging significantly improved contrast-to-noise ratio (CNR) by 124.37% compared to B-mode images.
• Benign tumors showed a significantly higher fuzzy entropy value (0.033) than malignant tumors (0.022), enabling classification.

Conclusions:

• The ultrasound fuzzy entropy breast imaging method enhances imaging performance and lesion detection capabilities.
• Fuzzy entropy measures microscopic tissue chaos, improving scattering information and lesion detectability.
• This method accurately classifies benign and malignant lesions by analyzing ultrasound signal causality.