Early Pancreatic Cancer Detection Using Deep Learning and Optical Neural Networks on Integrated Photonic Chips

Employing Integrated Photonic Chip-Based Optical Neural Networks (ONNs) for early pancreatic cancer detection, achieved an average Dice score of 0.58 for 2D and 0.6412 for 3D segmentation, matching or surpassing electronic models, demonstrating ONNs as efficient, high-speed alternatives to traditional electrical training systems.