Radiative transfer is a fundamental tool in astrophysics for understanding physical states and formation processes of various objects. In this talk, I will demonstrate how similar principles can be applied to a distant field—medical diagnostics using near-infrared light. Biological tissues are highly scattering media, analogous in some respects to dusty astrophysical systems, where photons undergo complex trajectories before reaching the observer. By adapting the techniques in the astrophysics research, we have developed a new radiative transfer code, TRINITY, that can simulate time-resolved photon transport in the human head and enables us to address the associated inverse problem. I will present our simulation results and their integration with machine learning, which enables rapid identification of bleeding sites. Our AI-based diagnostic model achieves an accuracy of over 90% in detecting bleeding sites. Finally, I will discuss ongoing efforts toward practical applications, including early-stage detection of brain hemorrhage, and future directions for interdisciplinary research connecting physics and medicine.