New Tissue Image Collection Advances Disease Research
The spora dataset standardizes thousands of images to help scientists improve AI analysis of tissue function and health.
Updated on Oct. 28, 2026 in Diseases — General
Live Poll
Should large-scale medical datasets be made open-access to accelerate artificial intelligence development?
Researchers at EPFL and the European Bioinformatics Institute have released spora, a large-scale open-access collection of 13,546 spatial proteomics images. This resource compiles data from 5,477 patients to help scientists standardize and compare research on how tissues function.
Why it matters
Spatial proteomics is crucial for understanding disease development, but existing datasets are often difficult to combine due to variations in equipment and storage. By standardizing these images, the new collection helps AI researchers more easily analyze tissue function across different studies.
The spora collection aggregates 13,546 multiplex images from 5,477 patients, spanning 14 tissue origins and five distinct imaging technologies. While the dataset provides a unified standard, the efficacy of using this specific collection to train future disease-diagnostic AI remains under investigation.
The players
EPFL
A Swiss public research university focusing on science and engineering, leading the development of the spora collection.
European Bioinformatics Institute
A global research organization focused on bioinformatics and data sharing that collaborated on the project.
The details
Spatial proteomics allows researchers to map the location of proteins within tissues, providing a view of how cells interact in health and disease. The team spent two years organizing thousands of images by standardizing protein annotations and matching medical information to specific cells. This process overcomes barriers created by differing laboratory equipment and storage protocols that previously hindered large-scale data integration.
Timeline
Scientists spent two years creating the spora collection between 2024 and 2026.
Health Landscape
This release follows the trend set by large-scale biological data initiatives like the Human Protein Atlas that prioritize open access to accelerate cross-disciplinary research. It marks a significant shift in spatial proteomics by providing a unified, interoperable resource.
While this tool is designed for researchers, the standardization of these tissue maps accelerates the discovery of how diseases develop at the cellular level. This work may eventually lead to more accurate diagnostic tools that your physician can use to interpret tissue samples in the future.
The takeaway
Accessing high-quality, standardized data is the primary hurdle for modern medical AI development in spatial proteomics. Researchers continue to expand the collection, and following updates to these large-scale open databases can provide insight into the changing trajectory of disease detection.
Further reading
Explore more developments in Diseases — General to understand how standardized data is shaping future clinical research.
Source note: This article includes information reported by Swiss Federal Institute of Technology, Lausanne (EPFL).
Live Poll
Should large-scale medical datasets be made open-access to accelerate artificial intelligence development?
