Supervisor: Dr. Vincent Detours
Context. Descriptions of tissues by histopathologists rest on verbal statements limited by quantitative inaccuracies and personal cognitive biases. We proposed an unsupervised computational framework that transposes to histology the concepts and methods of RNA-seq gene expression analysis, setting histology on an unbiased quantitative ground. Leveraging this framework and the GTEx dataset, we built an atlas surveying the histological diversity of 40 organs from 946 non-diseased individuals and documenting 10,056,047 associations between 2,560 morphemes—the histological analogs of genes in our framework—and 9 layers of patho-clinical and multi-omic molecular data, providing a rich context to interpret histology. In contrast with the ideally healthy normal specimens depicted in histology textbooks, the atlas reveals the influence of age, sex, genetics and sub-clinical pathologies on tissue structures.
An alpha version of our Atlas of human histological diversity can be browsed here:
The related preprint is available here: https://www.biorxiv.org/content/10.64898/2026.01.30.702249v1
Project. The successful applicant will:
- Introduce the concept of magnification/scale in the morphological analysis underlying the atlas.
- Generate automatic annotations of morphologies from vision/language LLM, and from datasets orthogonal to slide collections (e.g. clinical data, genomics, transcriptomics,…).
- Generate atlases from in-house and collaborators datasets.
Keyword. Artificial intelligence, LLM, foundation models, digital pathology, histopathology, whole slide images
Skills. The applicant has a master computer sciences, mathematics, physics or bioinformatics and has
- Outstanding coding skills
- A working knowledge of web development
- A working knowledge of AI tools, including proven fluency with pytorch
- Is eager to learn about biomedicine and follow medical school training in histology
- Is able to communicate effectively in an interdisciplinary environment
- Is able to multitask between projects in a highly dynamic research environment