Label-Free Detection of Acrosome Reaction of Human Sperm Based on Diffraction Phase Microscopy

The acrosome reaction (AR) is essential for successful fertilization, which is crucial for live sperm with fertilization potential in clinical practice. However, label-free assessment of the AR status of sperm is still challenging. In this study, a multimodal imaging system has been developed, integrating with fluorescence imaging, phase imaging, and bright-field imaging for observing the same sperm. The sperm phase characteristics were analyzed to assess the AR status. According to the phase in the acrosome and nucleus, the three-phase metrics ( $\stackrel{\circledR }{{\phi }_{acr}}$ , $\stackrel{\circledR }{\Delta \phi }$ , $\stackrel{\circledR }{R_{\phi }}$ ) are screened out to distinguish the AR status with comparison of fluorescence imaging, while $\stackrel{\circledR }{R_{\phi }}$ has the highest accuracy. Based on $\stackrel{\circledR }{R_{\phi }}$ , the classification model quantified by ROC analysis provided a high performance with an AUC value of 0.907. After excluding sperm head with vacuoles, the classification accuracy of $\stackrel{\circledR }{R_{\phi }}$ has been further improved to 95.5%. The predicted AR status for live sperm is close to that reported data. These results demonstrate that label-free diffractive phase microscopy can quantitatively provide the detection method of AR status for live sperm, which is a promising technical approach for clinical applications.