Reproduce “nnInteractive Agent: General-Purpose AI Agents as 3D Image Annotators” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Fabian Isensee, Constantin Ulrich Harsy, Maximilian Rokuss, Benjamin Hamm, Philipp Vollmuth, Klaus H. Maier-Hein
nninteractive-agent enables general-purpose AI agents to create 3D medical image masks through nnInteractive. The agent interprets a request, locates targets, places spatial prompts, inspects masks, and corrects errors, while nnInteractive generates the segmentation. On a 14-image benchmark, Codex’s unguided track scored mean Dice 0.842, versus 0.709 for a one-point baseline, though this small study includes no human comparison.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.