Enhancement of Digital Mammography Images Using Neutrosophic Divergence Score Based on Intuitionistic Fuzzy Entropy

Uncertainty in medical images-especially mammograms-caused by low contrast and insufficient brightness creates difficulties in detecting masses and microcalcifications. These limitations often lead to diagnostic uncertainty for radiologists, making effective image enhancement essential for accurate clinical assessment.

This study aims to develop an improved method for digital mammography image enhancement that reduces uncertainty, improves contrast, and preserves fine details to support more accurate diagnosis.

The proposed method integrates intuitionistic fuzzy entropy and neutrosophic sets (NSs) in a five-stage framework: (1) Transforming the input image into an intuitionistic fuzzy set; (2) Applying intuitionistic fuzzy entropy to reduce ambiguity; (3) Converting the image to an NS representation; (4) Enhancing image details using the neutrosophic divergence score (NDS); (5) Improving contrast through fuzzy histogram hyperbolization. Performance was evaluated on two benchmark mammography datasets using quantitative metrics, including the contrast improvement index, discrete entropy, absolute mean brightness coefficient, absolute mean brightness error, and the naturalness image quality evaluator, as well as a qualitative visual assessment.

Experimental results show that the proposed method surpasses existing approaches, including intuitionistic fuzzy sets, type-2 fuzzy sets, and neutrosophic-based enhancement methods. It achieves superior contrast enhancement, preserves naturalness, and effectively highlights fine mammographic details.

The method substantially reduces uncertainty in mammography images, enhancing diagnostic visibility and supporting improved accuracy for radiologists. Its strong performance across fatty, fatty-glandular, and dense-glandular breast tissues makes it a promising component for future computer-aided diagnosis systems.