Tumor and microcalcification characterization using Entropy, Fractal Dimension and intensity values statistical analysis in mammography

TL;DR

This study proposes using entropy and fractal dimension analysis to automatically segment and distinguish malignant tumors and microcalcifications in mammograms, improving diagnostic accuracy.

Contribution

The paper introduces a novel method combining entropy and fractal dimension for automatic segmentation and classification of mammographic abnormalities.

Findings

Fractal dimension is higher in tumor/microcalcification areas.

Entropy is lower in these areas due to uniform intensity.

The method accurately locates abnormal regions based on intensity analysis.

Abstract

Digital analysis of mammographic images is a complementary tool to clinical evaluation, commonly used to identify tumors and/or microcalcifications in mammograms. Recent mammographic equipment, can automatically classify them using this methodology. The difficulty in finding and classifying such areas, arise from different factors such as: image acquisition methodology, excess of brightness, similar physiological and radiological properties of tissues, etc. In this work it is proposed that the numerical computations of fractal dimension and entropy are tools that could be used to automatically segment and distinguish malignant tumors and/or microcalcifications in digital mammograms. The study consisted in segment the image in two areas: background and confirmed malignant tumor and/or microcalcification, to which the fractal dimension and entropy values are calculated and it was found the correlation between them. The analysis was performed on a grayscale re-digitized images from images provided locally by hospital for microcalcifications and from public databases for malignant tumors. From these re-digitized images, it was able to give the exact coordinates as the diagnosis provides. For any image, the highest intensity value was located in the tumor and/or microcalcification area, resulting that the fractal dimension had a higher value than the rest of the image, while the entropy value was lower, due to the uniformity of intensities in these areas. To complete this study, it was performed a data analysis with the set of intensities in each area. This allowed to distinguish between the areas of interest according to the value of the intensities, having a value greater than 3 {\sigma}. Finally, it is shown a technique to visually highlight the malignant tumor and/or microcalcification.