Red blood cells are essential for oxygen transport and immune function in the human body. When these cells become abnormally shaped, they can indicate serious health conditions, including diabetes, malaria, hereditary blood disorders, and vascular diseases. Currently, medical diagnosis relies primarily on optical microscopy, which requires fluorescent staining and manual inspection by trained experts. This traditional approach suffers from significant limitations: it is time-consuming, subjective, poorly reproducible, and often misses subtle changes in individual cells. Other advanced techniques like electron microscopy and super-resolution imaging can visualize cells at the nanoscale but are expensive, operationally complex, and still require staining. There is an urgent clinical need for rapid, objective, and label-free methods to analyze red blood cell deformations at the single-cell level.

A research team led by Prof. Nan Zeng at Tsinghua Shenzhen International Graduate School has now developed an innovative solution using polarized light technology. The method, called dual-angle Mueller matrix polarimetry (DMMP), analyzes individual blood cells by measuring how they alter the polarization state of light. The system shines polarized light at cells and captures their unique polarization signatures. By combining theoretical modeling with machine learning algorithms, the researchers extracted six specific polarization feature parameters that quantitatively describe cell size, shape, refractive index, and surface characteristics. The team validated their approach by testing red blood cells under various stress conditions that mimic disease states. Results showed that the polarization parameters could effectively distinguish between normal cells and those with abnormal deformations. Furthermore, when analyzing mixed blood samples containing both healthy and abnormal cells, a Random Forest classifier using these parameters achieved over 94% accuracy in determining cell type proportions. The technology offers several key advantages: it requires no staining or complex sample preparation, can analyze hundreds of cells per minute for high-throughput screening, and provides objective, quantitative measurements rather than subjective visual assessments. This label-free approach could be readily adapted for routine blood screenings in hospitals and clinics, potentially enabling earlier disease detection, more accurate diagnosis, and better monitoring of treatment responses for blood-related disorders.
The work entitled “Analysis of erythrocyte deformation characteristics based on dual-angle Mueller matrix measurement” was published in Frontiers of Optoelectronics (published on Dec. 3, 2025).

DOI:10.1007/s12200-025-00166-2