Theme

Announcing the MICCAI 2026 FIDO Challenge: Multimodal AI in Ophthalmic Surgery

April 7, 2026

•

Research

The organizers of the MICCAI 2026 FIDO Challenge

Ophthalmic surgery demands precise, real-time image guidance to navigate complex tool maneuvers and perform delicate tissue manipulation. To push the boundaries of surgical assistance, we are thrilled to announce the Fusion for Intelligent Decision-support in Ophthalmology (FIDO) Challenge, an official challenge of the International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI) 2026.

The Organizers & The Mission

Organized and funded by SynthesEyes GmbH and ZEISS (Carl Zeiss Meditec AG), the FIDO Challenge aims to drive innovation in multimodal AI-driven decision-support systems. Currently, surgical systems typically rely on either a wide-field fundus microscope image or high-resolution intraoperative Optical Coherence Tomography (iOCT) independently. To bridge this gap, SynthesEyes and Carl Zeiss Meditec AG are releasing a first-of-its-kind, highly specialized synthetic dataset to power real-time, multimodal data fusion.

Participants will tackle two primary tracks:

Task 1: Surgical Instrument Keypoints Spatial Localisation (Extending conventional 2D keypoint detection into pseudo-3D spatial localization and tool-tissue distance estimation leveraging both modalities).
Task 2: Intraoperative OCT to Fundus Registration (Accurate 3D-to-2D spatial alignment of volumetric iOCT data with 2D microscope fundus images).

A Thematic Day on Ophthalmology

The FIDO Challenge will be a part in the thematic day on Ophthalmology at MICCAI 2026. We are excited to hold the event in conjunction with the 13th Ophthalmic Medical Image Analysis Workshop (OMIA13) and GAVE challenge. OMIA13 will encompass a broad range of topics in ophthalmic image analysis, including retinal biomarker discovery, multimodal imaging, and the application of advanced computer vision and machine learning techniques. The workshop will also address key challenges such as improving accuracy, ensuring reliability and robust validation, and accounting for the wide diversity of ocular conditions.

Event Highlights

Beyond the competition, the challenge day is designed to foster deep, multidisciplinary collaboration. The event will feature:

Exclusive Keynotes by leading minds in the field, including Prof. Nassir Navab and expert retina surgeon Dr. Koorosh Faridpooya.
A Complementary Panel Discussion bringing together thought leaders across industry, academia, and clinical practice to debate the future of multimodal imaging and decision-support systems in the operating room.

Important Dates & Submissions

Dataset Release: Mid. May, 2026.
Submission Details: To be announced soon! (Model submission and automatic evaluation will be managed through Codabench).
Challenge Event: October 4/8, 2026.

Stay tuned for more updates on the official challenge website, to learn how you can participate and access the dataset.

Organizers