Dr. Eyal Zimlichman, founder and director of ARC Innovation, a global healthcare innovation testbed, and chief innovation and transformation officer at Israel’s Sheba Medical Center, is a leading voice in healthcare IT.
He sees provider-facing generative AI-powered systems expanding rapidly in 2025, automating administrative tasks to ease workforce shortages and burnout. Further, he believes patient-facing AI, like virtual nursing avatars, will progress more slowly due to regulations. But both, he said, will enhance accessibility and cut costs.
On another health IT front, he foresees progress for precision medicine this year. Digital pathology adoption will accelerate, delivering biomarker results in minutes instead of weeks, he said. AI integration with multi-omics – genetics, metabolomics, proteomics – will lead to tailored treatments and better outcomes, he predicted.
We interviewed Zimlichman to talk about AI and precision medicine and get him to offer advice to CIOs and other IT leaders at hospitals and health systems about what they should be keeping their eyes on in the coming year.
Q. Where do you see provider-facing AI heading in 2025?
A. Physician burnout and workforce shortages persist as significant challenges in healthcare, and generative AI can mitigate these issues by streamlining workflows and reducing administrative burdens.
In 2025, provider-facing AI will expand significantly to reduce workloads across health systems by automating tasks like data entry and transcriptions, reviewing EHRs and files to extract relevant patient data, and other solutions that will greatly increase efficiency and allow providers to focus on patient care over administrative tasks.
With development of specific use case foundation models, such as for radiology, digital pathology, etc., we will see clinical decision support at a completely new level, compared to what we used so far. I predict we will start seeing regulatory approvals, first for specific use cases, and later for foundation models. This will bring about slowly growing integration of these solutions to clinical files.
Finally, with regulation stepping in, and better definitions of the ethics and responsible use of AI, I predict professional associations to also rise to the challenge and start addressing AI technologies as an integral part of clinical guidelines. Building on emerging scientific evidence, this will bring about the beginning of an era where provider-facing AI will become common and an integral part of clinical care.
Q. What are your predictions for patient-facing AI in 2025?
A. Patients are directly suffering from workforce shortages, especially in fields like mental health that are sorely lacking from sufficient providers. That’s why I think we’re going to see a huge drive toward solutions where AI, and specifically generative AI, can help us confront this challenge.
In 2025, patient-facing AI will play an increasing role through systems like digital nursing assistants to provide information on how to handle different conditions or generative AI-driven mental health therapies that can provide first-line support, offering patients immediate care and assisting clinicians in managing larger patient volumes.
These platforms can streamline intake processes, enhance accessibility and reduce wait times, allowing professionals to treat more patients efficiently. While AI will support – not replace – clinicians, it will help improve care accessibility and alleviate pressure on the healthcare system, which will considerably reduce human workload and drive down costs.
I predict we will also start seeing the emergence of virtual health coaches, based on genAI, trained for specific purposes – from wellness and aging in place, all the way to chronic conditions such as heart failure and diabetes.
Virtual coaches provide the potential for lifestyle changes, including medication adherence, as well as allow empowering patients by providing the right information at the right time for them to make informed decisions.
Q. What do you see happening in the field of precision medicine in 2025?
A. There are current breakthroughs in precision medicine that have already begun, which I expect will expand and advance significantly in 2025 and beyond – particularly in the identification of biomarkers that enable more tailored treatments for individual patients.
We will likely see providers leveraging AI to analyze existing multi-omics data, extending beyond genetics to include areas like digital pathology, metabolomics and proteomics. These technologies offer deeper insights into how treatments can be customized based on a patient’s unique profile.
With digital pathology becoming more common, we will start seeing AI-powered digital pathology, which will allow for faster results in biomarker identification, enabling personalized treatment to begin the same day of diagnosis, rather than waiting weeks for genetic sequencing, which is the current norm.
This approach will allow for earlier intervention, improving patient outcomes and prognosis and transforming how healthcare is delivered.
Precision medicine will also allow us to tailor prevention based on individual predictions. Large data sets, including novel biomarkers such as microbiome, metabolomics and even vocal biomarkers, will allow us to predict specific diseases such as cardiovascular, cancer and dementia. Prevention efforts will no longer be a “one size fits all” approach.
Q. Overall, what should CIOs and other IT leaders at hospitals and health systems be keeping their eyes on in the coming year?
A. We are getting to a point where data infrastructure and standards allow organizations to better leverage their data assets. Big data platforms would drive better access to data analytics and insights across the organization.
Providing solutions that allow for self-service of staff members, together with the needed culture change, would unlock huge potential for organizations to become more effective and efficient. Each organization should build a strategic plan to get there and look for best in class platforms that will allow this transformation to take place.
With the fast development and expansion of genAI solutions, CIOs should be familiar with the existing solutions, mostly centered at this point at the most important aim of reducing burden on staff and addressing workforce shortages. CIOs should learn how to assess these solutions and the required regulation around AI.
The huge potential that is inherent within genAI, coupled with the risks that still exist, will require familiarity with this field. Integration into EHRs will make it easier to integrate, and we are likely to see a race of leading EHR vendors to develop/buy systems and integrate as a differentiator in this field.
Follow Bill’s HIT coverage on LinkedIn: Bill Siwicki
Email him: bsiwicki@himss.org
Healthcare IT News is a HIMSS Media publication