GenMC: In Vivo Photoacoustic and Ultrasound Dataset for Blood Oxygen Saturation Estimation

This dataset supports the manuscript “GenMC: Real-Time Generative Monte Carlo Surrogate for Quantitative Photoacoustic Imaging”. It contains in vivo data from 15 volunteers with different skin tones who were included in the quantitative blood oxygen saturation (sO₂) analysis. GenMC is a conditional generative framework trained on Monte Carlo (MC)-generated optical fluence maps. It predicts spatially resolved fluence from ultrasound-derived tissue geometry and assigned optical properties. The predicted maps are used for photoacoustic signal compensation and spectral unmixing to estimate blood oxygen saturation. Source code: https://github.com/MengjieSHI/GenMC-optical-fluence-synthesis