Google Cloud’s Aashima Gupta sees AI as a ‘collaborator’ in healthcare

Much of the information in health care records doesn’t appear in text.

Aashima Gupta, Google Cloud’s global director of healthcare strategy & solutions, says healthcare organizations are using AI to improve efficiency and reduce the stress of their workforce.

Google Cloud has been developing new AI-powered tools to help clinicians find the information they need in charts, images and X-rays. The company touted new generative AI capabilities in Vertex AI Search for healthcare at the HIMSS Global Health Conference & Exhibition in Las Vegas last week.

Aashima Gupta, Google Cloud’s global director of healthcare strategy & solutions, said the new improvements will allow a clinician to easily and quickly find out vital information that comes from charts or other visual images. A new “Visual Q&A” tool will also be able to analyze data in tables and provide relevant findings in searches.

Google Cloud also touted new advanced AI agents designed to help prepare patient charts and perform other administrative tasks. The company says Google’s Agentspace will make it easier for customers to build their own AI agents to do research more easily and to manage certain tasks. AI agents are designed to perform tasks autonomously.

In a conversation with Chief Healthcare Executive® at the HIMSS conference, Gupta says AI tools in the healthcare industry are maturing rapidly. She says healthcare organizations are able to adapt AI into more tasks to reduce administrative burdens and burnout of their workers.

“Healthcare used to be lagging behind on adoption of technology,” Gupta says. “This moment, in AI, is different. We are seeing Gen AI really help with some of that burnout, those administrative burdens.”

• Read more: Using AI to help doctors connect with patients

AI’s use in healthcare is surging. More than half of America’s hospitals are using AI predictive models to gauge which patients could be at greater risk of complications, according to a study published in Health Affairs in January. Two out of three doctors say they’re using AI in their practice, up from 38% in 2023, according to an American Medical Association survey.

Generative AI tools allow clinicians to quickly get summaries of research. Gupta says AI agents are going to be playing a bigger role. Basalt Health is using AI agents and other Google technologies to help medical assistants get ready for patient visits, she notes.

Hospitals can utilize AI agents to improve a number of administrative functions, including revenue cycle management, managing supply chains and scheduling surgical procedures.

“Agentic AI is able to plan for multiple steps ahead, create a goal, and be able to research it,” Gupta says. “So really, AI is a true collaborator.”

The emergence of “semantic AI” is also a noteworthy development, she says. Semantic AI tools aren’t bound by simply matching the keywords in searches. With semantic search, a clinician can look for shorthand, medical terminology and abbreviations and still be able to find the desired information.

Highmark has partnered with Google Cloud to use AI tools to improve management of claims submissions and authorization requests with Allegheny Health Network. Google Cloud has also collaborated with the Mayo Clinic to develop generative AI capabilities in clinical settings.

• Read more: Mayo Clinic’s John Halamka: ‘We have to use AI’

Hospitals and healthcare organizations that are looking to acquire AI tools should consider what problems they’re trying to solve. “Not all AI products are Gen AI products, not all Gen AI products are agentic AI products,” she says.

Organizations shouldn’t be fixated on the latest shiny object. They should figure out the right use cases and “solve for that vision,” she says.

Customers that have made the most progress with AI tools have sound approaches and governance in place to examine new solutions and determine how they should be used. Health systems need to get input from across the organization in choosing and using new AI tools, and they should have mechanisms to evaluate how they’re working.

“You need a solid technology strategy guiding that transformation. And we see that with some of our forward leading customers,” she says.

Gupta says she’s hopeful that AI can be used to improve health equity and close disparities in outcomes among underserved groups. AI tools can help flag patients who are overdue for a colorectal screening or a mammogram.

Some health leaders have been concerned about AI tools reflecting racial bias in health care, but Gupta says she’s optimistic that new technologies will help clinicians use the best information.

“If you think about health equity, there has been bias in the data for decades,” she says. “That data exists, but can AI be an assistive tool to help us understand bias? I see that as a promise of generative AI to help us be more cautious.”