To resolve the inherent trade-off between imaging depth and throughput in organoid phenotyping, we developed a fully integrated platform for longitudinal monitoring. This workflow spans the entire experimental pipeline: from the computer-aided design (CAD) of specialized hardware and the generation of diverse brain organoids to high-field 9.4T MRI acquisition and the construction of a foundational 4D database.
The hardware component incorporates a custom 15-well MRI-compatible plate featuring a specialized U-bottom geometry. This design gravitationally centers each organoid at the magnet's isocenter, ensuring maximized signal homogeneity and spatial standardization across all 15 samples imaged simultaneously.
Unlike optical imaging modalities limited by light scattering in mature, millimeter-scale tissues, our T2-weighted 3D RARE MRI sequence provides unrestricted depth penetration with an isotropic spatial resolution of approximately 40 μm. The platform captures full-volumetric structural contrast, revealing internal features such as ventricular cavities and tissue density gradients without the need for clearing or fixation.
Leveraging this high-throughput capability, we constructed FORMA (Four-dimensional Organoid Resonance Mapping Atlas), the largest longitudinal MRI dataset of human brain organoids to date. This open-source atlas comprises over 1,700 MRI volumes collected from three distinct brain regions to maximize morphological diversity:
The dataset also spans varying genotypes (2 healthy control and 2 SCZ patient-derived lines) and a broad developmental window. Furthermore, the non-invasive nature of our platform enables true longitudinal tracking, allowing us to map the region-specific morphological trajectories of individual organoids over months.
9.4T high-field MRI
~40 μm isotropic
T2-weighted 3D RARE
15 organoids per scan
FORMA Atlas is an open-source project, making the largest longitudinal MRI dataset of human brain organoids freely available to the research community. We encourage researchers to explore, analyze, and build upon this foundational dataset.