In the Global IndiaAI summit a session titled, 'GPAI Convening on Global Health and AI', was organized. Global Health and AI focuses on healthcare concerns in global South Asian countries and gathers insights on AI utilization. The session discussed the integration of AI into health care, mainly focusing on the unique challenges and opportunities it presents. 

The session focused on data quality and accessibility, digital infrastructure, and regulatory and legal frameworks. These issues are particularly relevant in the Global South, where high-quality, reliable data and robust digital infrastructure are often lacking. The discussion aimed to understand Global viewpoints, learn from best practices, and explore how these can be adapted to improve implementation in low-resource settings.

The speakers for the session were Dr Kartik Adapa, Regional Adviser, Digital Health, WHO-SEAR; Dr Basant Garg, Addl CEO at NHA; Dr Madhukar Bhagat, Joint Secretary Ministry of Health & Family Welfare; Dr Mona Duggal, Associate Professor of Community at Post Graduate Institute of Medical Education and Research Chandigarh and adjunct professor at Indraprastha Institute of Information Technology, Delhi, and IIT Ropar, Mr Mihir Kulkarni, ML scientist at Wadhwani AI, Mr Sameer Kanwar, Director, Digital Health with PATH (India & South Asia Hub), Dr Sanjay Sarin, VP of Access and Regional Programmes, FIND, Ms Megha Dada Chawdhry, Advisor at BrainSightAI amongst other distinguished speakers. 

Dr Kartik Adapa said, “The Global South faces challenges in using AI, which can be addressed by using a generalized AI life cycle, embedding that with some core principles for trustworthy AI, and developing a framework for identifying the challenges in using AI in healthcare. The eventual goal is to create an AI playbook with the help of stakeholders.”

The challenges that Southern nations are facing in terms of AI usage in Global Health are:

• Addressing local priorities
• Data scarcity and quality
• Infrastructure limitations
• Contextual differences
• Ethical concerns
• Workforce capacity
• Regulatory framework
• Cost and sustainability
• Cultural acceptance
• Interoperability 

Dr Madhukar Bhagat discussed the need for a robust regulatory and legal framework for AI in healthcare. He also shared various AI solutions, such as Clinical Decision Support System (e Sanjeevani), Media Disease Surveillance (IHIP), and Diabetic Retinopathy (AIIMS Delhi, Rishikesh), which are already deployed, and ones like Dermatology & Cataract (AIIMS Delhi and Rishikesh), Oral Cancer & Alcoholic Hepatitis (PGI Chandigarh), and ASHA AI (NHSRC), which are under deployment in various geographies across India. 

The session also discussed how it is important to validate the available data in building a framework that all the stakeholders trust. The need for an institutional framework in the form of Centers of Excellence (CoE) that could certify the quality and efficiency of AI models will be a significant step forward for AI in healthcare.

Dr. Basant Garg during his thought sharing said “We as thinkers and stakeholders around AI need to create a regulatory framework of policies which is equally applicable to both private and Govt sectors” 

The panellists discussed how solutions to various challenges require the right policies around usability and efficacy, fairness and equity, safety and reliability, transparency and accountability, and security and privacy. Dr Adapa also said, “Education and capacity building around AI is crucial. We must move from a virtual and siloed education system to a more diverse and holistic perspective.”

Ms Megha Dada, who also serves as an advisor at BrainsightAI, said, “We are creating Google Maps for the brain, for which we need specific and contextual data sets. For us, the question of availability would greatly depend on joint intervention and how we get more investment for India-based data sets.”

Mr Mihir Kulkarni talked about data curation for models working in healthcare: “If a model goes wrong and breaks, we need to evaluate ‘why did it go wrong?’ To understand the problem deeply, we need domain experts, public health experts, and doctors.”

He added, “High-quality curated data with humans in the loop is better than curating it on large amounts of data where the quality of the data might be uncertain.”  

Dr Mona Duggal affirmed, "In the healthcare segment, regulatory frameworks must be adaptable to specific cities and not follow other geographies. Ethical committees at medical colleges need to have interoperability and must have data scientists with a high level of sensitivity. Even in rural areas, we need awareness of consent and data.”

The speakers concluded the session by stating that India is an excellent platform for launching healthcare in AI. We need availability, affordability, and accessibility when setting healthcare goals and integrating AI.