Microtubule Segmentation Paper Published in PLOS Computational Biology

Our paper Synthetic data enables human-grade microtubule analysis with foundation models for segmentation has been published in PLOS Computational Biology.

Researchers from HTW Berlin, BHT and the Max Planck Institute for Infection Biology worked on it together. Machine learning experts brought foundation models and a synthetic data pipeline, and biologists brought the microscopy data and the questions that matter in the lab. It is a good example of what Appl-FM is built for: foundation model research that answers real questions in the sciences.

The project page explains how the synthetic data is generated and what the SynthMT benchmark shows.