Advanced Computational Frameworks for High-Resolution Biospeckle Image Analysis: Integrating Machine Learning and Optical Coherence Techniques for Enhanced Biological Tissue Characterization

This work provides a novel computer vision technique for high-resolution biospherical imaging based on cutting-edge machine learning and optical coherence technology. Our approach improves the accuracy and ease of identification of living cells when compared to previous methods. We conducted extensive ablation research on the framework’s components. This demonstrated how different deep learning layers pick up distinct information and how optical coherence facilitates image comprehension. Systems that operate continuously and in real time, using minimal computational resources, can use the proposed approach. The study discovered that the structure is robust and useful for researching biological tissues. Furthermore, a hyperparameter sensitivity study indicated that the model required fine-tuning to function more effectively. This novel design expands the scope of biological tissue analysis and establishes the foundation for future advancement by offering a comprehensive and adaptable solution for high-resolution biospeckle imaging.