Revolutionizing early breast cancer screening: Advanced multi-spectral transmission imaging classification with...

Breast cancer stands as a formidable danger to the health of women worldwide, underscoring the critical need for effective screening methods. Multispectral transmission imaging offers a promising avenue due to its non-invasive potential for early screening. Some researchers already suggested registration to solve the problem of jitters due to respiration and movement and frame accumulation technology to solve the low grayscale problem. However, the classification of blood vessels and breast tissue in breast images often suffers from low signal-to-noise ratio (SNR) and low contrast, hindering accurate classification. This paper proposes a novel improved Otsu's method with K-Means clustering to address this challenge. The proposed method aims to enhance the foundation for classification using multispectral transmission images. The study utilizes multispectral transmission images captured at four wavelengths, representing an innovative avenue for early, affordable breast cancer screening research. Initially, 300 images are registered and accumulated to prove the efficiency of the suggested methodology. Then, median filtering is applied to reduce noise in the images. Improved Otsu's segmentation method is then employed to separate blood vessels from breast tissue. After that, K-means clustering is utilized to accurately classify these components. The results of the proposed method demonstrate significant improvements in classification accuracy and grayscale contrast of multispectral breast images. By effectively distinguishing blood vessels and breast tissue, the proposed methodology addresses the inherent challenges of low contrast in multispectral transmission imaging. This advancement offers a clearer pathway for early breast cancer screening.