Accuracy assessment of tissue polarimetric properties based on Mueller matrix images decomposition

Abstract

Mueller matrix images(MMI) contain complete polarization information of the media. Mueller matrix decomposition technique, where Mueller matrix polar decomposition(MMPD) and differential decomposition(MMDD) are widely used to decompose MMI, is the key to extract intrinsic polarimetry characteristics of biological tissues. For the decomposition of biological tissue MMI, Satish et al. expressed that MMDD was more suitable for Mueller matrix polarimetric analysis of tissues, while Alali et al. pointed out that MMDD did not offer a great advantage over MMPD. To deal with this problem, we explore how to choose the appropriate decomposition method to accurately extract the polarization information in biological tissues. The experimental results indicate that the linear retardance and optical rotation images obtained from two decomposition methods are different if tissues exhibit significant linear retardance and optical rotation effects simultaneously. According to the physical model of MMDD that the occurrence of polarization effects is orderindependent, MMDD should be applied to MMI of tissues to obtain accurate polarization characteristics in this situation. The biological tissue has low optical rotation in most cases in which the polarimetric images extracted from two decomposition methods are nearly identical, so MMPD and MMDD both can accurately acquire the polarimetric properties of tissues. Meanwhile, comparing the runtime of two decomposition methods to process MMI, we find the processing speed of MMDD is much faster than MMPD. Thus, we summarize that MMDD method is more suitable for the decomposition of the biological tissue MMI, with the advantages of both fast and accurate, which is significant in diagnosis of clinical.