O-PRESS: Boosting OCT axial resolution with Prior guidance, Recurrence, and Equivariant Self-Supervision

Kaiyan Li, Jingyuan Yang, Wenxuan Liang, Xingde Li, Chenxi Zhang, Lulu Chen, Chan Wu, Xiao Zhang, Zhiyan Xu, Yueling Wang, Lihui Meng, Yue Zhang, Youxin Chen, S. Kevin Zhou

Abstract: Highlights•A prior guided, robust method enhances OCT axial resolution in real-time.•Self-Supervised learning achieves performance akin to fully-supervised.•Deep learning incorporates the OCT physical model.•Zero-shot learning seamlessly extends to diverse tissues and imaging systems.•Achieve higher resolution with existing hardware sources.