Data associated with the publication: CODAvision: best practices and a user-friendly interface for rapid, customizable segmentation of medical images

The provided dataset includes a ‘Sample Dataset’ folder with four whole-slide images (WSIs) of murine lung histology in .ndpi format, each paired with a manual annotation file in .xml format. These pairs are used to train an image segmentation convolutional neural network (CNN) with the CODAvision software. An additional WSI is provided for testing, along with its corresponding annotation file located in a subfolder named ‘test’.

A second folder is also included ‘Processed sample dataset’, containing the processed output generated after the dataset has been run through CODAvision. This folder provides the metadata produced during processing, the performance report, and the output files generated by the trained model.

A third folder, ‘Additional annotated images.zip,’ contains annotated images used for the snapshots and workflow figures presented in the manuscript.

Lastly, users may train a model and segment the provided images by following the steps described in the accompanying ‘DEMO CODAvsion.pdf’ file.

Users may download the CODAvision software by following the instructions provided in the CODAvision GitHub repository:<a href="https://github.com/Kiemen-Lab/CODAvision">https://github.com/Kiemen-Lab/CODAvision</a>. They can then use the ‘DEMO CODAvision.pdf’ file to parametrize the dataset and run the processing steps. A preservation copy of CODAvision software version as of 2026-04-29 is included with the download files but users should get the latest version at the GitHub repository link.