Analysis of polarization features of human breast cancer tissue by Mueller matrix visualization

Thi-Thu-Hien Pham,Thao-Ngan Ngoc Quach,Quoc-Hoang-Quyen Vo
DOI: https://doi.org/10.1117/1.jbo.29.5.052917
IF: 3.758
2024-01-15
Journal of Biomedical Optics
Abstract:SignificanceBreast cancer ranks second in the world in terms of the number of women diagnosed. Effective methods for its early-stage detection are critical for facilitating timely intervention and lowering the mortality rate.AimPolarimetry provides much useful information on the structural properties of breast cancer tissue samples and is a valuable diagnostic tool. The present study classifies human breast tissue samples as healthy or cancerous utilizing a surface-illuminated backscatter polarization imaging technique.ApproachThe viability of the proposed approach is demonstrated using 95 breast tissue samples, including 35 healthy samples, 20 benign cancer samples, 20 grade-2 malignant samples, and 20 grade-3 malignant samples.ResultsThe observation results reveal that element m23 in the Mueller matrix of the healthy samples has a deeper color and greater intensity than that in the breast cancer samples. Conversely, element m32 shows a lighter color and reduced intensity. Finally, element m44 has a darker color in the healthy samples than in the cancer samples. The analysis of variance test results and frequency distribution histograms confirm that elements m23, m32, and m44 provide an effective means of detecting and classifying human breast tissue samples.ConclusionsOverall, the results indicate that surface-illuminated backscatter polarization imaging has significant potential as an assistive tool for breast cancer diagnosis and classification.
radiology, nuclear medicine & medical imaging,biochemical research methods,optics
What problem does this paper attempt to address?