Light-RepViTSR: Ultra-Lightweight Super-Resolution for Real-Time Photoacoustic Endoscopy in Tumor Biopsy

Guanyi JIANG¹Rui Yang²Zhanfeng FANG²Yuwei LUO²Xianghu YU²Li LIU^2*

• ¹Peking University Shenzhen Hospital Department of Hematology, Shenzhen, China
• ²Great Bay University, Dongguan, China

Real-time in situ biopsy marks a paradigm shift in clinical oncology by enabling immediate intraprocedural pathological diagnosis during endoscopy. Photoacoustic endoscopy (PAE) serves as a pivotal technology in this field, uniquely visualizing tumor microvasculature and hypoxia through synergistic fusion of optical contrast and ultrasonic resolution. However, PAE's inherent resolution-speed tradeoff in raster scanning induces severe motion artifacts from physiological activities (e.g., peristalsis, respiration), critically compromising diagnostic reliability. Although deep learning-based super-resolution (SR) techniques show promise for photoacoustic microscopy, their clinical translation to PAE is hindered by excessive computational demands and insufficient real-time performance. To overcome this limitation, we propose Light-RepViTSR, an ultra-lightweight SR reconstruction network based on the RepViT architecture and specifically optimized for real-time PAE. Our approach integrates the representational capacity of RepViT's re-parameterizable convolutional blocks while eliminating non-essential components (e.g., squeeze-and-excitation layers) to maximize computational efficiency. Comprehensive evaluation on a multi-source dataset—including 19 previously unseen murine cerebrovascular images and 18 self-collected plant vein images—demonstrates Light-RepViTSR's superiority. The network consistently outperforms conventional methods across scaling factors (×2, ×4, ×8), achieving significant improvements in PSNR (up to +1.41 dB at ×8) and SSIM (up to +0.047 at ×2) while reducing model size by > 99% and inference time by > 60% versus SRResNet. This work establishes a pathway toward practical real-time high-resolution PAE, demonstrating significant potential to enhance in situ tumor biopsy accuracy.