Abstract: Human intestinal organoids (HIOs) are invaluable in regenerative medicine and disease modeling due to their ability to mimic the architecture and function of the human intestine. The use of HIOs is growing rapidly, driven by advancements in stem cell-derived organoid technologies. A key challenge in studying and utilizing HIOs is the non-invasive monitoring of their development and cellular status. Most existing imaging techniques rely on genetic or molecular labeling, which can disrupt organoid integrity and require extended sample preparation times.
In this study, we demonstrate label-free monitoring of three-dimensional (3D) subcellular structures in HIOs using holotomography (HT) enhanced by artificial intelligence (AI). HT is a label-free imaging method for measuring the refractive index (RI) of cells and has recently proven effective for imaging organoids. By combining HT with AI, we fully exploit HT’s potential to profile HIOs at the subcellular level. This integration overcomes the molecular ambiguity of RI; the AI translates HT images of HIOs into fluorescence-equivalent images of subcellular structures, including nuclei and cell membranes, typically achieved through labeling. Our results show that AI-assisted HT enables virtual staining of organoids without any genetic or molecular intervention, providing a safe and efficient way to monitor the temporal dynamics of HIOs in applications such as disease modeling and drug screening.
Funding Source: This work was supported by grants from NRF Korea (RS-2024-00442348, 2015R1A3A2066550, 2022M3H4A1A02074314), IITP (2021-0-00272), MOTIE/KIAT (P0028463), and KFRM (21A0101L1-12).
