NIH Scientists Develop "Digital Twin" Of Eye Cells to Understand and Treat Age-Related Macular Degeneration

Why It Matters

The digital twin provides a scalable, high‑resolution framework for AMD drug discovery and can be extended to any condition where cell polarity drives pathology, accelerating therapeutic pipelines across biomedicine.

Key Takeaways

• First subcellular-resolution digital twin of human primary cell
• AI model POLARIS segments 3‑D RPE structures automatically
• Atlas maps polarized vs non‑polarized RPE states
• Enables rapid screening of AMD therapeutic targets
• Platform adaptable to other eye and non‑eye diseases

Pulse Analysis

The emergence of a digital twin for retinal pigment epithelial cells marks a turning point in cellular modeling. By integrating 1.3 million high‑content confocal images with a custom AI pipeline, researchers achieved subcellular resolution that captures nucleus position, organelle morphology, and apical‑basal polarity. This level of detail, previously attainable only through labor‑intensive microscopy, now resides in a computational atlas that can be queried, simulated, and iterated at scale, opening new avenues for precision biology.

Age‑related macular degeneration hinges on the loss of RPE polarity, which disrupts photoreceptor support and triggers vision loss. The POLARIS‑generated atlas delineates the trajectory from immature to fully polarized RPE cells, highlighting structural checkpoints that fail in disease. Pharmaceutical teams can overlay candidate compounds onto this framework to predict how interventions restore polarity or prevent degeneration, dramatically shortening the pre‑clinical validation cycle. Moreover, the model supplies a quantitative baseline for biomarker discovery, enabling clinicians to monitor disease progression with unprecedented granularity.

Beyond ophthalmology, the digital twin concept is poised to reshape research on any tissue where cell orientation matters, such as kidney tubules, intestinal epithelium, and cancer metastasis. By coupling AI‑driven segmentation with mathematical modeling, scientists can generate virtual organelles for diverse cell types, fostering cross‑disciplinary collaborations and reducing reliance on animal models. As the biotech industry embraces these in‑silico platforms, investment in high‑throughput imaging and AI infrastructure is likely to surge, accelerating the translation of cellular insights into marketable therapies.

NIH scientists develop "digital twin" of eye cells to understand and treat age-related macular degeneration

Tuesday, February 10, 2026

National Institutes of Health (NIH) researchers have developed a digital replica of crucial eye cells, providing a new tool for studying how the cells organize themselves when they are healthy and affected by diseases. The platform opens a new door for therapeutic discovery for blinding diseases such as age‑related macular degeneration (AMD), a leading cause of vision loss in people over 50.

“This work represents the first ever subcellular resolution digital twin of a differentiated human primary cell, demonstrating how the eye is an ideal proving ground for developing methods that could be used more generally in biomedical research,” said Kapil Bharti, Ph.D., scientific director at the NIH’s National Eye Institute (NEI).

The researchers created a highly detailed, 3‑D data‑driven digital twin of retinal pigment epithelial (RPE) cells, which perform vital recycling and supportive roles to light‑sensing photoreceptors in the retina. In diseases such as AMD, RPE cells die, which eventually leads to the death of photoreceptor cells, causing loss of vision.

For RPE cells to do their multiple jobs properly, they require a top‑to‑bottom polarity: the cell’s “top” (apical) side faces photoreceptors, where they recycle worn‑out photoreceptor parts daily. The cell’s “bottom” (basal) side faces the blood supply, bringing in nutrients and oxygen and shipping out waste.

Researchers constructed the digital twin based on RPE cells made at NEI from induced pluripotent stem cells (iPS) developed by the Allen Institute for Cell Science, Seattle. 3‑D imaging data for 1.3 million RPE cells, taken from nearly 4,000 fields of view, were collected using an automated confocal microscope.

Using the imaging data, the researchers trained an artificial‑intelligence (AI) algorithm they called POLARIS (polarity organization with learning‑based analysis for RPE image segmentation) to recognize the nucleus and other cell structures, as well as the cell’s shape and volume. 3‑D segmentation data (labels assigned to image voxels) were generated over different stages of cell development.

The team paid particular attention to polarity, quantifying the size and shape of the cell, its organelles and cytoskeletal structures, including 3‑D spatial localization at various stages of development. They found that healthy, developing RPE cells follow a predictable path toward a polarized state.

The resulting AI‑driven atlas of polarized and non‑polarized RPE cells provides researchers with a reference for studying how diseases affect RPE at the cellular and subcellular levels, which could be transformative for therapeutic discovery.

“The digital twin approach represents a powerful new tool for AMD therapeutic development and could be adapted to study other eye and non‑eye diseases and conditions affecting cell polarity,” said Bharti.

“By combining AI with mathematical modeling, we’ve created a window into cellular processes that were previously hidden from view,” said the study’s first and senior author, Davide Ortolan, Ph.D., NEI research fellow. “This technology doesn’t just help us understand what’s happening in AMD, it gives us a platform to discover how to fix it.”

This research was funded by the NIH/NEI Intramural Research Program.

---

Reference

Ortolan D, Sathe P, Volkov A, Reichert D, Sebastian S, Maminishkis A, Schaub NJ, Ljungquist B, Bose D, Ferrari J, Lin N, Pegoraro G, Simon CG, Sharma R, Bajcsy P, and Bharti K. “AI driven 3D subcellular RPE map discovers cell state transitions in establishment of apical‑basal polarity.” Published February 6, 2026 in Nature Partner Journal‑AI. https://doi.org/10.1038/s44387-026-00074-6