Transformation of full 4 × 4 Mueller matrices: a quantitative technique for biomedical diagnosis

Polarization images contain abundant microstructural information of samples. Recently, as a comprehensive description of the structural and optical properties of complex media, the Mueller matrix imaging has been widely applied to biomedical studies, especially cancer detections. In previous works, we proposed a technique to transform the backscattering 3 × 3 Mueller matrices into a group of quantitative parameters with clear relationships to specific microstructures. In this paper, we extend this transformation method to full 4 × 4 Mueller matrices of both the back and forward scattering directions. Using the experimental results of phantoms and Monte Carlo simulation based on the sphere-cylinder birefringence model, we fit the Mueller matrix elements to trigonometric curves in polar coordinates and obtain a new set of transformation parameters, which can be expressed as analytical functions of 16 Mueller matrix elements. Both the experimental and simulated results demonstrate that the transformation parameters have simple relationships to the characteristic microstructural properties, including the densities and orientations of fibrous structures, the sizes of the scatterers, and the depolarization power of the samples. We also apply the transformation parameters of full 4 × 4 Mueller matrices to human liver cancerous tissues. Preliminary imaging results show that the parameters can quantitatively reflect the formation of fibrous birefringent tissues accompanying the cancerous processes. The findings presented in this study can be useful for in vivo or in vitro polarization imaging of tissues for diagnostic applications.