In a significant step forward for healthcare innovation in India, Eka Care has announced the public release of sizable, anonymized healthcare datasets. This initiative marks a pivotal moment for the Indian healthcare AI ecosystem, providing researchers and developers with invaluable resources to build homegrown AI models specifically tailored to the Indian population's healthcare needs.
Researchers and developers interested in accessing these datasets can find them on the IndiaAI’s AIkosh platform, where they are available for download under appropriate usage terms. Additionally, these datasets are also available on HuggingFace.
One of the biggest challenges in developing HealthCare AI in India is the lack of large-scale, high-quality healthcare datasets that represent the unique characteristics of the Indian population. Most AI models currently deployed in Indian healthcare settings are trained on datasets from Western populations, potentially limiting their efficacy and accuracy when applied to Indian patients.
Eka Care's release aims to address this fundamental gap by providing access to meticulously curated, anonymized healthcare data from Indian subjects.
The release includes two distinct datasets, each offering unique insights into different aspects of healthcare:
1. NidaanKosh: With more than 6.8 million laboratory readings from 100,000 Indian subjects, this dataset offers unprecedented insights into diagnostic patterns and biomarker distributions in the Indian population. Researchers can leverage this extensive collection to develop more accurate diagnostic algorithms and reference ranges specifically calibrated for Indian patients. AIKosh | HuggingFace
2. Spandan: This unique dataset features over 1 million Photoplethysmography (PPG) signals from Indian subjects. PPG technology, which measures blood volume changes in the microvascular bed of tissue, is increasingly used in wearable devices for monitoring heart rate and other vital signs. This dataset will enable developers to create more accurate algorithms for cardiovascular monitoring devices used by Indian populations. AIKosh | HuggingFace
By releasing these datasets, Eka Care is empowering the Indian AI community to develop healthcare solutions that address the specific challenges and requirements of the Indian healthcare system. IndiaAI is hosting the datasets on the AIkosh platform, making them easily accessible to researchers, startups, and established organisations alike.
We believe that by opening access to this valuable resource, we can catalyze a new wave of innovation in Indian healthcare AI.
It's important to note that all datasets have been rigorously anonymized to protect patient privacy while maintaining their utility for research and development. Eka Care ensures that no personally identifiable information is included in the released data.
This release is expected to accelerate the development of AI models catering to Indian Healthcare needs. With access to these datasets, researchers can now train algorithms that account for the genetic, environmental, and lifestyle factors specific to Indian populations.
This initiative by Eka Care represents a remarkable step toward democratizing healthcare AI development in India and fostering solutions that are truly built for India, by India.
While the volume of the shared datasets is significant, it represents a fraction of healthcare data in Eka Care. We leverage the entire data to create purpose driven small LLMs for specific healthcare needs. A documentation of the MedAI services offered by Eka Care can be found here.
.jpg)
