Characterization of Biological Tissue Structures Using Poincare Sphere

Objective In recent years, the incidence of cancer has been increasing annually and it has become one of the diseases with the highest fatality rate. The development of modern medicine has put forward high requirements for the early diagnosis of diseases. At present, for the detection of biological tissues, methods such as X-ray imaging, magnetic resonance imaging, and ultrasound imaging are primarily used. However, these detection methods can be used only as auxiliary methods and not as a basis for final diagnosis. Pathological examination remains the current "gold standard" for cancer detection. According to the data reported by the 2020 World Congress of Pathology, the global digital pathology market has been growing every year since 2016. However, with the rapid growth of the pathology market's demand, problems such as an acute shortage of pathologists, low automation of pathology departments, and long diagnosis time remain. Therefore, it is important to develop a fast and effective pathological examination method. Methods Compared with the traditional nonpolarized optical detection methods, Stokes-Mueller matrix imaging technology reflects rich biological tissue microstructure information and is widely used in tissue lesion detection. This study explores the Poincare sphere characterization method of biological tissue structure. This method employs the Mueller matrix polarization decomposition method to extract three characterization vectors to characterize the basic polarization characteristics of the medium, i. e. , retardance vector (R), diattenuation vector (D), and polarizance vector (P), draw the P and D vectors on the Poincare sphere, and compare the vector difference with the center distance of the sample points as the evaluation index. Taking the nucleus and fiber structure of skeletal muscle tissue and fibrous connective tissue as the objects, this study explores the distribution law of the representation vectors of different structures of the same biological tissue on the Poincare sphere and further compares and analyzes the distribution law of the representation vectors of the similar structures of different biological tissues on the Poincare sphere. Results and Discussions We use the fibrous connective tissue to explore the mapping of different magnifications of the same tissue on the Poincare sphere. The images of the P and D vectors at 9.2 x and 18 x , respectively, are drawn on the Poincare sphere (Figs. 4-7) . The figures reveal that the spatial distribution of the two color points is very close but they show aggregation. When the polarization staining method is used to characterize different structures of biological tissues and the diattenuation values of two structures are approximately equal, the pseudocolor map of diattenuation values shows that the whole image tends to be of the same color and the boundary between the muscle fiber and nucleus is fuzzy and difficult to distinguish. Using Poincare spheres, the differences between structures can be compared from three dimensions, thus improving the resolution compared with polarization staining method. The P and D vectors of the skeletal muscle and fibrous connective tissues at an 18 x ratio are used to compare the mapping of similar structures of different tissues on the Poincare sphere. The distribution of the P vector in the skeletal muscle tissue is similar to that in the fibrous connective tissue. The D vectors of the skeletal muscle and fibrous connective tissues are distributed in the same quadrant of the Poincare sphere with similar specific positions [Figs. 7 (a) and 8 (b) . Because the skeletal muscle and fibrous connective tissues show similar fibrous structures, the distribution of the P and D vectors on the Poincare sphere conforms to our expected results. The difference of the P and D vectors between the two tissues is determined, and the difference for the fibrous connective tissue is >8% than that of the skeletal muscle tissue (Table 1). Conclusions Accurate detection of different structures in biological tissues is of great significance for clinical applications. In this study, a fully polarized Mueller matrix imaging detector for biological tissues is constructed to investigate the distribution of different representation vectors of the same sample or similar structures of different tissues on the Poincare sphere. In the current study, the spherical and column scatterers represented by nucleus and fiber are studied. The characterization effect of the P and D vectors at a high magnification is found to be better than that at a low magnification. In terms of the difference of the P and D vectors between the two tissues, the difference for the fibrous connective tissue is >8% than that of the skeletal muscle tissue. In general, the P vector is generally 19.7% better than the D vector for obtaining the difference between two structures in a single biological tissue. Internal biological tissue microstructure changes tend to be small and difficult to detect. The experiment conducted in this study confirms that the Poincare sphere can be used to distinguish different structures of biological tissues; the minimum surrounded by different scatterer balls can realize effective determination of biological tissue microstructure and has certain clinical application potential.