Photo: MoMo Productions/Getty Images
Google Cloud showcased new artificial intelligence tools and AI agents this week. The company says they could help healthcare and other organizations unlock new returns on investment and reshape how healthcare is delivered, accessed and improved.
"In healthcare, the top use cases are in tech support (53%), security operations and cybersecurity (49%), productivity and research (46%) and patient experience (44%)," said Aashima Gupta, Google Cloud's global director of healthcare strategy and solutions, and Shweta Maniar, its global director of life science strategy and solutions.
Several healthcare organizations are showcasing how they're putting the tools to work. New Jersey-based Hackensack Meridian Health announced that it is using Google AI through integrations with its electronic health records, while California-based Color has launched an AI assistant to automate breast cancer screening eligibility and scheduling. New York-based Castor also introduced the Catalyst AI platform to automate patient enrollment and consent processes in clinical trials.
Improving flows to benefit patients
Patient experiences are inextricably linked to both the speed and quality of care, and healthcare consumers want to see healthcare rapidly adopt digital patient engagement and data-driven decision-making.
"The ongoing transition to value-based care models emphasizes patient outcomes and cost-efficiency, making digital tools that empower patients to actively participate in their own care even more essential," David Nickelson, senior partner and consulting lead, health and life science, Randstad Digital, told Healthcare IT News in January.
Hackensack Meridien, like many health systems, is using AI to deploy clinical note summarization.
Relying on Google Gemini, 12 specialties now have access to genAI capabilities in their Epic Systems EHR, according to Sameer Sethi, the health system's senior vice president and chief AI officer.
"With the help of Google, we try to build these capabilities and then integrate into clinical workflow," he said during a media roundtable hosted by Google on Tuesday. "It's as simple as framing that right within Epic," he explained, noting that the health system is also building prediction models through an integration into its EHR.
The health system said it is not only reducing the time specialty staff spend in EHR workflows by between 5% and 20%, but it has also built an AI-powered lab summarization agent that enables patients to get their test results faster and understand them better.
The AI agent summarizes lab panel results, highlights significant trends and key findings, and generates preventive care recommendations for primary care physicians to use when drafting messages to patients about their lab results.
The tool speeds up communications to patients and leads to faster follow-up and timelier preventive actions, the health system said in a statement on Thursday.
The third way Hackensack Meridian said it is using AI is by immediately equipping its Neonatal Intensive Care Unit nurses with knowledge, regardless of their current workload or experience level.
With rapid access to the most current NICU best practices and policies through the NICU Nurse Agent, the health system said it aims to ensure that the highest standards of care are met in the unit.
While it is looking to AI to stop the erosion of clinician well-being and the pervasive discontinuity of patient journeys, Hackensack Meridian said its overall AI strategy is creating a scalable, replicable model for value-based care.
The approach moves beyond "incremental optimizations," and the strategic AI deployments address healthcare's "most entrenched systemic crises," Gupta said in the statement.
"They are establishing the blueprint for the next generation of [VBC]."
Democratizing patient access
California-based Color developed a new AI platform on Google Cloud Vertex AI to democratize patient care through its Virtual Cancer Clinic.
With the launch of Color Assistant on Thursday, the company said it is targeting the 20–30% of eligible women in the United States who are not up to date on their mammograms in order to tackle the concerning breast cancer metrics for women under the age of 50.
Diagnosis rates have increased by nearly 20% since the early 2000s among that population.
"We know how to stop late-stage breast cancer diagnoses through screening, but we need to make getting that screening much easier,” said Color CEO and cofounder Othman Laraki in the announcement.
The assistant determines eligibility for mammograms, schedules breast cancer screenings and coordinates follow-up, making them more widely and easily accessible to all patients through an initiative that runs through Dec. 31.
An automated bot answers questions and sets up mammograms through integrations with EHR systems, Laraki explained on Tuesday at the media briefing.
The virtual assistant, powered by the Gemini 2.5 family of models, also maintains clinical oversight, making it possible to serve a general population, he said.
The agentic framework relies on the American Cancer Society and other evidence-based guidelines. But the assistant requests clinical reviews by clinicians within Color’s 50-state medical group, and its care teams connect with those eligible for a mammogram for any necessary clarifications and to coordinate appointments. They can also order other types of imaging in accordance with clinical guidelines, the company said.
To close the loop and improve patient access, Color's clinicians receive and deliver screening results directly to patients and then coordinate next steps for abnormal findings with the patients’ existing care providers.
The new agent also checks transcripts to ensure quality results, Laraki noted.
"There are some parts of healthcare where AI can truly make abundance," he said.
Patient enrollment in clinical trials
Castor, a clinical trial technology company, released its Catalyst AI-powered platform developed on Google Cloud to automate tasks in clinical studies on Thursday.
The "self-driving" platform automates processes and compresses overall timeframes for administrative tasks, with its data-entry and verification skills launched through patient-mediated retrieval pathways.
That includes speeding up the rates of patient enrollment and consent through digital integrations with EHRs. The agent requests and retrieves patient consent using the HL7 FHIR standard, the company said in its announcement.
The AI platform and its event-driven data infrastructure, which the company said took 18 months to build, are designed to provide full observability, auditability and regulatory readiness for clinical studies.
"Without a detailed understanding of what's actually happening in a study, you simply can't apply AI effectively," Derk Arts, Castor's CEO, said in a statement.
Andrea Fox is senior editor of Healthcare IT News.
Email: afox@himss.org
Healthcare IT News is a HIMSS Media publication.
