by Pathology, the branch of medicine that deals with the study of diseases at the cellular and tissue level, has long relied on two-dimensional (2D) tissue slices for diagnosis. However, the intricate three-dimensional (3D) nature of human tissue poses a limitation to this traditional approach. To address this challenge, researchers from Mass General Brigham, in collaboration with the University of Washington, have developed Tripath, a deep learning model that can analyze 3D pathology datasets to make clinical outcome predictions.
The team imaged curated prostate cancer specimens using two advanced 3D high-resolution imaging techniques. Tripath was then trained to predict Prostate Cancer recurrence risk on volumetric human tissue biopsies. By capturing the 3D morphologies from the entire tissue volume, Tripath demonstrated superior performance compared to pathologists and outperformed deep learning models that rely on 2D morphology and thin tissue slices.
The findings, published in Cell, represent a significant step forward in the field of pathology. Although the new approach requires further validation in larger datasets before it can be implemented in clinical settings, the researchers are optimistic about its potential to revolutionize the way clinical decisions are made.
“Our approach highlights the importance of analyzing the whole volume of a tissue sample for accurate patient risk prediction, which is the hallmark of the models we developed and only possible with the 3D pathology paradigm,” said lead author Andrew H. Song, Ph.D., from the Division of Computational Pathology in the Department of Pathology at Mass General Brigham.
*Note:
1. Source: Coherent Market Insights, Public sources, Desk research
2. We have leveraged AI tools to mine information and compile it
