Rajroshan Sawhney: Harnessing AI for Healthcare in India
Rajroshan Sawhney plays a pivotal role at Google, leading initiatives related to Frontier and Applied AI Partnerships in the Asia-Pacific region, as well as spearheading efforts with the Health AI Developer Foundations (HAI-DEF). His work focuses on turning research into real-world applications, particularly in healthcare, where partnerships are helping to conduct over 10 million AI-driven screenings for critical illnesses such as cancer and tuberculosis.
In his position at Google, Sawhney has played a key part in the successful launch of MedGemma, a project aimed at improving healthcare delivery by leveraging AI. Before joining Google, he led strategic partnerships at a notable computer vision startup in San Francisco and Singapore. He has a degree in Computer Science from BITS-Pilani and is a member of Mensa International, as well as the India Young Leaders Forum.
In a recent interview, Sawhney discussed how Google is integrating AI into India’s healthcare sector to improve services for citizens.
Q. Google has rolled out AI solutions for diabetes, TB, breast cancer, and maternal health in India. Which area has shown the most significant results, and how do you measure success?
“Our main goal is to make advanced AI technology accessible to every level of healthcare,” Sawhney explained. “We do this by equipping healthcare providers with innovative tools and working with communities and governments to establish a strong foundation for widespread health initiatives.”
He mentioned that significant progress has been made in three areas: improving access to screenings with partners like Apollo Radiology International, enhancing maternal healthcare through predictive AI tools, and easing the administrative workload for healthcare workers.
Sawhney also emphasized the belief that solutions developed for India could be beneficial globally. For instance, their diabetic retinopathy model has been successfully tested in partnership with local institutions and is now helping facilitate screenings for over 600,000 people worldwide.
Q. Transitioning AI from pilot projects to national implementation can be challenging. What obstacles do you face in India, and how is Google addressing them?
“Scaling AI in healthcare requires collaboration among various stakeholders, including healthcare providers, policymakers, and tech companies,” Sawhney noted. He highlighted the importance of creating solutions that can be integrated with existing systems to address real challenges faced by healthcare professionals.
He also pointed out efforts like AIIMS using MedGemma models, which cater to India’s diverse languages and cultural contexts while maintaining safe and secure systems without the need for expensive hardware.
Q. Google is collaborating with the National Health Authority on converting unstructured medical records into a standardized format. How transformative will this be for India’s health system, and what privacy concerns exist?
Sawhney expressed excitement about the ongoing transformation, particularly with the new Aarogya Setu 2.0 app, designed to streamline individual health profiles using standardized data formats. He emphasized the importance of data privacy at every step, ensuring users maintain control over their health information.
Q. Why is it crucial for India to develop healthcare AI models specific to local languages and diseases?
Healthcare is deeply personal and varies greatly across different regions. As such, it’s vital that AI models are tailored to meet local needs rather than relying solely on international models. An example illustrated was the development of IndusDerma, focused on improving dermatological care for individuals with diverse skin tones, thus enhancing accuracy and reducing biases found in traditional datasets.
Q. Looking ahead, which area of healthcare do you believe could see the biggest breakthroughs in the next few years?
Sawhney believes that AI will play a crucial role in transforming healthcare delivery and accelerating medical research. He highlighted that AI is already being used effectively for early screenings and diagnostics, and he anticipates significant advancements as these technologies become more accessible to local communities.
Q. With concerns around AI’s accuracy and patient privacy, how does Google ensure its AI models can be trusted, especially in a diverse country like India?
Google collaborates closely with partners to ensure that patient data is handled securely and kept private. Sawhney stated that Google maintains a strong focus on building trust, highlighting rigorous training protocols on diverse medical data to develop safe and effective healthcare applications.
Overall, Sawhney’s initiatives represent a significant stride toward a more efficient and accessible healthcare system in India, showcasing the potential of AI to make a global impact.
