Medical Robotics Foundation Models

Surgical world models, Open-H-Embodiment, and computer vision for robotic training workflows.

My medical robotics work applies computer vision and generative models to surgical training, robotic manipulation, and future foundation models for healthcare robotics.

Open-H-Embodiment dataset overview

The public work spans diffusion-based suturing world models, Open-H-Embodiment contributions, surgical phase recognition, and computer vision scoring for endoscopy training. Together, these projects show how deployed perception and generative modeling can support training, assessment, simulation, and eventually more capable medical robots.

Role: Technical project lead or contributing author across medical AI projects with Columbia and Northwell Health collaborators
Keywords: Video diffusion, surgical action modeling, foundation model datasets, automated skill assessment, surgical workflow understanding, medical robotics

Links: Suturing GitHub, Open-H GitHub, Hugging Face model, Suturing World Models, suturing arXiv, Open-H project, Open-H arXiv.

References

2026

Open-H-Embodiment: A Large-Scale Dataset for Enabling Foundation Models in Medical Robotics

Nigel Nelson, Juo-Tung Chen, Jesse Haworth, Xinhao Chen, Lukas Zbinden, Dianye Huang, Mattia Ballo, Filippo Filicori, Mehmet Kerem Turkcan, and others

Apr 2026

arXiv DOI HTML Code Dataset Model

Open-H-Embodiment is a multi-institution open dataset for medical robot learning with synchronized video and kinematics across healthcare robotic platforms, enabling foundation models and world models for medical robotics.

2025

Towards Suturing World Models: Learning Predictive Models for Robotic Surgical Tasks

Mehmet Kerem Turkcan, Mattia Ballo, Filippo Filicori, and Zoran Kostic

2025

arXiv DOI HTML Code Model

This work fine-tunes video diffusion models on annotated surgical suturing demonstrations to model ideal and non-ideal fine-grained robotic surgical actions.