2023 has been dubbed as generative AI’s “breakout” year and tech giant Google is flexing its muscles once again in the healthcare space.
At the HLTH 2023 conference this week, Google Cloud announced updates to its Vertex AI Search engine for healthcare and life sciences organisations.
The new feature enables medically tuned generative AI powered searches on data including clinical notes and can be combined with Google’s large language model system Med-PaLM 2 so clinicians can find answers to medical questions directly from a patient's medical records. The system can also assist with answers to general medical questions.
"Bringing Google-quality, gen AI search capabilities across an organization’s entire ecosystem, including EHRs, has the potential to dramatically improve efficiencies, provide clinical decision support, and increase the quality-of-care clinicians can give patients," said Burak Gokturk, VP and general manager, Cloud AI and Industry Solutions for Google Cloud.
"Making Vertex AI Search more useful for healthcare and life science organizations is a priority for us because we know that having the right information and insights at the right time can make all the difference in health."
Data privacy concerns persist
Speaking in the opening session on Monday 9 October, Google Senior Vice President of Research, Technology & Society, James Manyika said protection of personal and privacy continues to be “fundamentally important” to Google and one of the biggest lessons learned from recent company controversies.
“Google Cloud doesn’t capture any private customer data,” stressed Manyika. “Also, when it comes to advertising we are trying to be thoughtful about making sure we don't provide advertisements targeted to people based on their health conditions. These things are enshrined in our principles.”
But as technology develops– the questions around data privacy are also shifting and evolving, he said. “We are constantly doing research to make sure we are protecting privacy. However, in the age of AI there is often a presumption that personal data is always needed for technology advances. Often the key is just analysing patterns in data and images.”
Manyika highlighted areas like cancer treatment where patients are constantly being screened. “For example in radiotherapy, doctors are constantly looking at CT scans in a process called contouring delineation and it can take hours for them just to understand how to separate the cancerous cells out of the organs from the healthy cells, but you don't need the patient's personal data to do this,” he said.
“It is all about analysing images and by using AI techniques we can enable physicians and care providers to be able to do the work of contouring and take it down from doing it manually in seven hours to a few minutes digitally.”
