What You Should Know:
– PathAI, a leader in the field of AI-powered pathology, has announced the launch of Pathology Universal Transformer (PLUTO), a foundation model specifically designed for the intricacies of pathology, promising to transform the development of AI tools for biopharmaceutical companies and pathology labs.
– In the realm of machine learning, foundation models (FMs) play a crucial role. These models are trained on vast amounts of unlabeled data using a technique called self-supervised learning. This process allows them to capture essential information from the data, forming a strong base for tackling specific tasks in diverse contexts.
PLUTO’s Advantages: Efficiency & Performance
Leveraging a high-quality FM like PLUTO offers significant advantages. Compared to building custom tools from scratch, adapting PLUTO for specific tasks reduces development time and costs. For instance, a single adaptation can be used for tasks like cell and nucleus segmentation on microscopic immunohistochemistry (IHC) slides, or histological subtyping on whole-slide H&E slides.
Developing effective FMs in pathology is particularly challenging due to the vast array of tasks involved, ranging from cellular analysis to whole-slide image (WSI) evaluation. PLUTO addresses this complexity by being specifically designed for the diverse needs of pathology.
Building a Superior AI Foundation
PathAI trained PLUTO on a massive dataset of unlabeled image patches, encompassing:
- Hundreds of millions of image patches
- Approximately 160,000 WSIs
- Over 30 disease areas
- Numerous indications, stains, scanner types, and magnifications
- Data from over 50 sources
To ensure PLUTO learns high-quality representations from this data, PathAI researchers implemented a customized self-supervised training process. This process utilizes multi-scale, flexible deep vision transformer architectures with innovative training objectives.
PLUTO’s Superior Performance
Rigorous evaluation has confirmed PLUTO’s capabilities. When adapted to various pathology tasks, PLUTO outperforms existing state-of-the-art models in areas like whole-slide analysis, cellular tasks, and subcellular tasks.
Furthermore, PLUTO empowers the development of AI-powered pathology tools at various scales and resolutions. Being more compact than previous models translates to significant reductions in training and inference costs. This efficiency enables highly scalable data generation and product development.
The Future of AI-Powered Pathology
PLUTO and its adaptations are poised to power the next generation of PathAI’s industry-leading suite of AI products for biopharmaceutical companies and diagnostic pathology labs. PLUTO’s capabilities extend beyond simply enhancing existing products. Its embeddings will pave the way for:
- Predicting underlying molecular alterations driving cancer and disease phenotypes, even those previously undetected by AI.
- Discovery of new biomarkers and therapeutic targets.
“PLUTO is meticulously designed to fuel research efforts for better diagnostic and prognostic tools across multiple disease areas. This technology will enable users to compress their pathology workflows using an exceptionally diverse, multi-resolution foundation model at significantly greater speed and a much broader scale than before,” said Harsha Pokalla, Head of Machine Learning at PathAI. “Our aim is to deploy this single, robust model to address a diverse range of applications and use cases within pathology and to continue developing new high performance AI tools that serve our mission to improve patient outcomes.”
