The National Health Authority (NHA) and IIT Kanpur have inked an MoU for development of digital public goods for Artificial Intelligence in healthcare.
Under this MoU, a federated learning platform across a variety of machine learning model pipelines, a quality-preserving database, an open benchmarking platform for comparing and validating AI models and a consent management system for research under the Ayushman Bharat Digital Mission (ABDM) would be developed by IIT Kanpur.
The platform will subsequently be operated and governed by the NHA, thereby unlocking the immense potential of AI for improving health outcomes, the Union Health Ministry said in a statement.
Speaking on the occasion, Union Health Secretary Apurva Chandra complimented the NHA and IIT Kanpur for signing "this very important MoU under the Ayushman Bharat Digital Mission that will provide an open public benchmarking platform for comparing and validating AI models".
Highlighting the importance of this MoU, Chandra stated that "use of data for improving health outcomes is the envisaged goal of this partnership". "It will use the data available under the ABDM to create a public benchmark for AI models for quantifying and diagnosing diseases against which other AI models can be benchmarked," he added.
Chandra said "one of the biggest issues related to AI models in the healthcare sector is the availability and efficacy of disease diagnosis data which will be tackled by this collaboration".
"In a clinical setting, the availability of reliable data will lead to better outcomes and result in better diagnosis," he added.
IIT Kanpur Director Professor Manindra Agrawal said "this partnership with the NHA will contribute towards democratizing access to quality healthcare in India".
"Partnering with the NHA, IIT Kanpur aims to create a transformative platform that will empower researchers, healthcare providers and policymakers to unlock the full potential of AI for improving health outcomes in India," he added.
This platform will offer numerous benefits by enabling providers of AI models for healthcare applications to perform genuine out-of-set validation and establish publicly verifiable performance benchmarks. This platform will foster trust in the consumer market for these applications.
The platform will address the fragmentation of health data in our country, making it easier for researchers to access and analyze health data without data leaving the custody of data fiduciaries, the statement said.
Healthcare data is uniquely costly to acquire and through this platform will be protected against statistical dredging, without compromising on openness and innovation, it added.
Published - September 11, 2024 05:30 pm IST
