AI-generated synthetic neurons speed up brain mapping

Research Google
Google Research developed the MoGen model to create synthetic neurons, reducing human proofreading labor in brain mapping by improving AI reconstruction accuracy.

Summary

Google Research has introduced MoGen, a generative AI model that creates realistic synthetic neurons to improve brain mapping accuracy. By training the existing PATHFINDER model with synthetic data, researchers achieved a 4.4% reduction in reconstruction errors. While seemingly modest, this efficiency gain could save approximately 157 person-years of manual proofreading for a full mouse brain mapping project. The team has released MoGen as an open-source tool to support ongoing efforts in connectomics.

(Source:Research Google)