Abstract
SIGNIFICANCE: Polarimetry offers advantages such as high information dimensionality and sensitivity to microstructures. Determining the depth of the tissue is essential for clinical diagnosis and treatment, such as lesion localization, removal, and drug delivery. However, relying solely on polarization techniques for tissue depth measurement remains a subject for further investigation. AIM: We aim to investigate the tissue depth measurement in turbid media using Mueller matrix polarimetry, with a focus on fibrous tissues. APPROACH: Tissue phantoms are constructed to quantitatively simulate fibrosis at specific depth. By analyzing Mueller matrix measurements across depth gradients, correlations between polarization basic parameters (PBPs) and tissue depth are established using supervised machine learning algorithms. RESULTS: We introduce an approach by combining degree of polarization (DOP)-sensitive PBPs with anisotropy-sensitive PBPs to develop depth-sensitive polarization feature parameters (DSPFPs). The DSPFPs exhibit enhanced sensitivity to depth in shallow layers while preserving accuracy in deeper layers. The effectiveness and robustness of the proposed method are validated through 2D depth-resolved imaging of tissue phantoms. CONCLUSIONS: We preliminarily explore the feasibility of depth measurement using Mueller matrix polarimetry, establishing a method for tissue depth assessment while also expanding the applications of polarimetry.