What is it about?
Computational framework to integrate imaging data of tissues generated from multiple imaging modalities and domains. Integration of imaging data measuring distinct feature spaces, on platforms that differ in resolution, information density, and data structure is a hurdle to harnessing highly detailed and valuable imaging data.. Our aim was to develop a platform and domain agnostic tool to integrate imaging data into one map tractable for downstream interrogation.
Featured Image
Photo by National Cancer Institute on Unsplash
Why is it important?
In the context of spatial biology, aligning data and annotations from histopathological stains, high-plex immunohistochemistry, and label-free detection by mass spectrometry imaging, is a growing area to develop comprehensive profiles and identify biomarkers from clinical tissues. Our approach keeps the human in the loop, enabling investigators from across disciplines to integrate their imaging data using an open source, explainable approach that can be used directly to identify correlates to groups or individual samples, as well as to generate robust training data.
Perspectives
MIAAIM was designed from the outset to ensure that when integrating complex, high-value imaging datasets, whether for biological discovery, to interrogate non-biological images, or to feed the training of models, the alignment underpinning that integration is mathematically rigorous, explainable, reproducible.
Patrick Reeves
Massachusetts General Hospital
Read the Original
This page is a summary of: MIAAIM: Multi-omics image integration with dimensional reduction for tissue state mapping, PLoS Computational Biology, May 2026, PLOS,
DOI: 10.1371/journal.pcbi.1014274.
You can read the full text:
Resources
Contributors
The following have contributed to this page
