Data and Intermediate Results from "ActiveVisium: Leveraging Active Learning to Enhance Manual Pathologist Annotation in 10x Visium Spatial Transcriptomics Experiments"

Description This dataset accompanies the publication “ActiveVisium: Leveraging Active Learning to Enhance Manual Pathologist Annotation in 10x Visium Spatial Transcriptomics Experiments,” presented at ECML PKDD 2025 (Applied Data Science Track). The dataset is organised into four main folders: Pathologist Annotations Contains manual pathologist annotations at the spot level for each sample used in the study. Data Includes all data used in the main experiments. Refer to the included README file for a detailed explanation of the folder structure. Annotation Consistency Experiment Data Provides data and results from annotation consistency experiments conducted on the CytAssist 11mm FFPE Human Kidney spatial transcriptomics dataset. These experiments identify potential noise in ground truth spot-level annotations (see Section 4.3 in the paper). More details are available in the corresponding README file. Test Data (Breast Cancer Sample) Contains a breast cancer tissue section, including: Histological image Spatial gene expression matrix Sample annotation file Useful for testing installation and validating the active learning workflow. _________________ Intended Use Reproduction of all results from the associated publication Benchmarking and development of new annotation or active learning strategies Testing and validation of the ActiveVisium software installation Licensing and Model Usage The ActiveVisium code and dataset are released under the GNU General Public License v3.0 (GPL-3.0). Note: Some foundational models used to generate features are subject to their own licenses and terms of use. Users are responsible for reviewing and complying with all relevant third-party licenses.