EndoLINA @ MICCAI 2026 — Endoluminal Intervention Navigation and Autonomy

About the Workshop

EndoLINA is a MICCAI workshop on endoluminal intervention navigation and autonomy with a deliberate focus on medical image computing (MIC) and computer-assisted intervention (CAI). Endoluminal procedures (e.g., bronchoscopy, colonoscopy, upper-GI endoscopy, ureteroscopy) operate in tortuous, deformable lumens with limited field-of-view and frequent visual degradation from motion, fluid/occlusion, specularities, and device/illumination artifacts. These conditions make navigation primarily an imaging-and-AI challenge: robust perception, multimodal fusion, uncertainty estimation, and clinically meaningful evaluation.

The workshop will highlight methods that connect pre-/intra-operative imaging to navigation decisions, covering anatomy-aware navigation representations, artifact-resilient visual localization, multimodal fusion and registration, uncertainty-aware decision-making, and benchmarks aligned with clinical workflows. The half-day format combines in-person keynotes, contributed talks, posters, and a panel to consolidate open problems and propose community benchmarks for trustworthy endoluminal navigation and autonomy.

EndoLINA is jointly organized with the ATM26 Challenge (Airway Tree Modeling), which involves both preoperative planning and intra-operative navigation for endobronchial intervention.

Topics of Interest

Navigation Representations

Anatomy- and topology-aware navigation representations and patient-specific map building from CT/MRI.

Visual Localization & Odometry

Artifact-resilient scope pose estimation from endoscopic video, including self- and weakly-supervised learning and domain generalization across devices/sites.

Multimodal Fusion & Registration

Fusion and registration of CT/MRI with real-time endoscopy (and fluorescence/US when applicable) for path planning and guidance.

Uncertainty-aware Decision Making

Long-horizon navigation with risk-aware policies and human-in-the-loop shared autonomy framed as CAI.

Benchmarks & Datasets

Metrics aligned with clinical workflow: target reachability, localization error, time-to-target, robustness to artifacts, failure prediction, and calibration.

Endoluminal Procedures

Applications in bronchoscopy, colonoscopy, upper-GI endoscopy, ureteroscopy, and other endoluminal interventions.

Important Dates

All deadlines are at 23:59 Pacific Time (PT)

Paper Submission Deadline

TBD

Notification of Acceptance

TBD

Camera-ready Submission

TBD

Workshop Date

October 4 or 8, 2026

Preliminary Program

Half-day schedule (200 minutes)

Time	Event
First Part (90 min)
08:00 - 08:05	Opening & Scope Framing
08:05 - 08:25	Keynote 1 — Prof. Nassir Navab (Technical Expert)
08:25 - 08:45	Keynote 2 — Prof. Alberto Arezzo (Clinical Expert)
08:45 - 09:30	Contributed Oral Session (3 presentations, 15 min each)
09:30 - 09:50	Break & Poster Session
Second Part (90 min)
09:50 - 10:05	ATM26 Challenge Overview
10:05 - 10:50	ATM26 Top-3 Team Presentations (15 min each)
10:50 - 11:10	Panel: "Benchmarks, Uncertainty, and Trustworthy Endoluminal CAI"
11:10 - 11:20	Closing & Next-step Roadmap

Keynote Speakers

Confirmed in-person speakers

Prof. Nassir Navab

Technical Expert

Technische Universität München
MICCAI Fellow

Prof. Alberto Arezzo

Clinical Expert

University of Torino

Additional keynote speakers from medical image computing, multimodal fusion/registration, and CAI with clinical deployment experience will be announced.

Call for Papers

Submission Guidelines

We invite high-quality original submissions that address challenges in endoluminal intervention navigation and autonomy. Topics include but are not limited to:

Anatomy- and topology-aware navigation representations
Artifact-resilient visual localization and scope pose estimation
Multimodal fusion and registration for path planning and guidance
Uncertainty-aware decision-making and shared autonomy
Benchmarks, datasets, and clinically aligned evaluation metrics
Self-supervised and weakly-supervised learning for endoscopic vision
Domain generalization across devices and clinical sites

Format

Papers should be formatted in LNCS style (Springer) and limited to 8 pages (plus references). Supplementary material may be submitted optionally. All submissions will undergo double-blind peer review with at least two reviewers per paper.

Proceedings

Accepted papers will be published in the MICCAI Satellite Events joint LNCS proceedings (Springer Nature). Top papers will be invited to submit extended versions to IEEE Transactions on Medical Robotics and Bionics and other related journals.

Submit Your Paper (Coming Soon)

Organizing Committee

General Co-Chairs

Alberto Arezzo

University of Torino

Guang-Zhong Yang

Shanghai Jiao Tong University
MICCAI Fellow

Nassir Navab

Technische Universität München
MICCAI Fellow

Organization Committee

Yun Gu

Shanghai Jiao Tong University

Elena De Momi

Politecnico di Milano

Yao Guo