File photograph used for representational purposes only
In an effort towards enhancing clinical research in India, the International Institute of Information Technology Hyderabad (IIIT-H) has joined hands with the Nizam’s Institute of Medical Sciences (NIMS) to unveil publicly accessible datasets featuring digitised histopathological images. These datasets include high-resolution images of brain cancer and kidney disease (Lupus Nephritis), forming the cornerstone of the India Pathology Dataset (IPD) project.
Histopathological images are microscopic images of tissues that are used to diagnose diseases, especially cancer.
The initiative offers benefits ranging from reduced risk of damaging physical slides to improved clinical decision-making and bettering research opportunities with the help of Artificial Intelligence (AI).
“Traditionally, tissue samples are examined under a microscope. By digitising these slides, computers can visualise the images, enabling collaborative diagnosis among pathologists across locations,” explained Vinod P.K., who leads the project’s dataset curation efforts.
The IPD initiative represents a multi-stakeholder collaboration among academia, hospitals, industry, and the government, aiming to digitise tissue biopsy slides. Supported by the Technological Innovation Hub for Data Banks, Data Services, and Data Analytics (TiH-Data), IIIT-H has installed a whole-slide digital scanner at NIMS to facilitate this effort.
Brain tumor research
One of the first datasets that has been released is the IPD-Brain dataset in Nature Scientific Data, an open-access, online-only journal for descriptions of scientifically valuable datasets.
Highlighting its clinical relevance, Megha Uppin from the Department of Pathology at NIMS noted that precise typing, sub-typing, and grading of cancers was critical for effective management. “This dataset allows machine learning models to explore regional and ethnic variations in brain tumours, enhancing diagnostic precision and identifying cancer subtypes,” she said.
Dr. Megha underscored the role of AI in bridging the gap created by the increasing complexity of molecular diagnostics and the shortage of specialised neuropathologists in India.
The team is also exploring AI models to predict molecular markers directly from tissue morphology, bypassing traditional and often expensive genetic lab testing. One such effort focuses on predicting IDH mutations, which are vital for diagnosing and prognosing brain tumours.
Expanding scope
The IPD project is expanding its datasets to include various cancers, such as breast, lung, colorectal, oral, and cervical cancers. NIMS is also contributing to a dataset on lung cancer. In addition to cancer research, the project has compiled a dataset on Lupus Nephritis, an autoimmune kidney disease prevalent among women in India, particularly in Telangana.
“This is one of the first instances of open-source medical data from India for the greater good,” said Mr. Vinod.
A second digital scanner has been set up on the IIIT-H campus, accessible to dental colleges, corporate hospitals, and other institutions for collaborative use. In the educational domain, these datasets provide medical students, especially pathology residents, an in-depth understanding of histopathology through high-quality digital images.
Published – January 22, 2025 12:15 pm IST
