fastER: a user-friendly tool for ultrafast and robust cell segmentation in large-scale microscopy

Area of Science:

• Cellular imaging and analysis
• Bioimage informatics
• Computational biology

Background:

• Quantitative cell microscopy generates vast datasets requiring automated analysis.
• Automated cell segmentation is challenging due to high cell density, variability, and low signal-to-noise ratios.

Purpose of the Study:

• To evaluate existing state-of-the-art cell segmentation tools for accuracy and speed.
• To develop a novel, fast, and accurate tool for large-scale cell segmentation in microscopy images.

Main Methods:

• Evaluation of multiple automated cell segmentation algorithms on diverse datasets.
• Development of fastER, a trainable cell segmentation tool with an interactive training interface.
• Application of fastER to segment and count cells in over 200,000 brightfield microscopy images.

Main Results:

• Existing tools showed variable performance and were often computationally expensive, limiting their use in large-scale studies.
• fastER achieved state-of-the-art segmentation quality at significantly higher speeds.
• fastER processed over 46 million cells from 200,000 images in approximately 2.5 hours on a desktop computer.

Conclusions:

• fastER provides a robust, efficient, and user-friendly solution for large-scale cell segmentation in microscopy.
• The tool's speed and accuracy facilitate high-throughput biological research.
• fastER is adaptable to various cell types and imaging modalities, making it broadly applicable.