Google’s vision for digital health: Leveraging AI for groundbreaking personal health insights

Google Health is setting a new benchmark in healthcare technology with its ambitious roadmap for 2024, focusing on the transformative potential of generative AI in enhancing clinical decision-making, understanding laboratory and imaging data, detecting diseases at an early stage, and offering customised health coaching based on user Fitbit data.

During the annual The Check Up event at Google’s Pier 57 office in Manhattan, key executives from the tech behemoth unveiled significant progress and updates on a range of leading-edge AI initiatives. The spotlight was on refining the Gemini model for healthcare applications, creating a bespoke health-centric large language model (LLM) for the Fitbit app to provide tailored wellness and health features, pioneering AI models for early disease detection, and exploring how generative AI can facilitate medical reasoning and support in clinical dialogues.

At last year’s Check Up event, Google introduced Med-PaLM 2, an LLM specifically tuned for healthcare. Following this, in December, Google announced MedLM, a suite of foundational models for healthcare built upon Med-PaLM 2, which has been made more widely accessible via the Google Cloud’s Vertex AI platform.

Various healthcare organisations are currently piloting Google’s LLM models to devise solutions across a spectrum of applications, such as enhancing clinical documentation and optimising nurse handovers.

Karen DeSalvo, M.D., Google’s Chief Health Officer, underscored the pivotal moment AI technology is within the healthcare sector during her address at The Check Up event.

“It’s an exhilarating time to be at Google, where our cutting-edge technology is leveraged to foster better health outcomes. Our mission is to embed AI in a manner that enriches people’s lives through improved health. We aim to achieve this by integrating health-focused solutions into everyday products and services, thereby empowering our partners and communities,” DeSalvo expressed.

DeSalvo shared her vision of a near future where AI empowers healthcare, enhancing the wellbeing of all, rather than a select few. However, she emphasised that AI is to be seen as a tool augmenting the clinician-patient interaction, rather than replacing it.

In collaboration with Google Research, Fitbit is developing a health-specific LLM to provide Fitbit and Pixel device users with a bespoke health coaching experience based on their personal data.

This initiative is set to offer personalised coaching functionalities, such as customised advice and recommendations, tailored to individual health and fitness aspirations. The launch of Fitbit Labs in the autumn marked the beginning of leveraging AI for deeper health insights.

Florence Thng, Director of Product Management at Fitbit, highlighted the collaborative efforts with Google Research and health experts to formulate personalised health LLMs that can interpret health and fitness data to offer customised advice, akin to a personal coach.

Moreover, earlier this year, Google unveiled AMIE (Articulate Medical Intelligence Explorer), an AI research system optimised for diagnostic reasoning and clinical dialogues, built on an LLM.

Greg Corrado, a distinguished scientist and the head of health AI at Google Research, during the event, emphasised the importance of empathy in medicine and how AMIE is designed to reflect this in clinical consultations.

The ambition to harness AI in healthcare is reaching new heights with Google’s ongoing research and development efforts. The Gemini model suite, particularly fine-tuned for the medical field, is demonstrating remarkable capabilities in advanced reasoning and handling complex data across multiple modalities.

The deployment of MedLM for Chest X-ray analysis signifies Google’s commitment to applying AI in practical, impactful ways in medicine, promising to assist healthcare professionals in making more informed decisions and improving patient care outcomes.

This evolving landscape of medical AI at Google points to a future where technology and healthcare intersect more seamlessly, promising a new era of innovation and enhanced care delivery.