AI in a White Coat
The generative AI revolution is transforming healthcare: from clinical documentation, through smart hospital management, to Epidemiological Management. And that’s not all; AI is already passing certification exams, diagnosing diseases and, in some cases, outperforming human doctors.
“We’re about three years into the generative AI revolution, and we’re already seeing dramatic changes in the healthcare system,” says Bruno Lavi, VP of Health at Matrix. “I believe that, in a few years, everything will look different. Systems are evolving, and AI solutions are entering every area, from clinical diagnosis to back-office and financial processes, and even performing actual medical procedures. And no less important is the fact that we humans – doctors, staff, patients – are all learning to operate in this new world.”
Regulation Playing Catch Up
The pace of change is rapid. As healthcare organizations worldwide adopt AI to improve the quality of care, cut costs, and reduce operational burdens, regulators are trying to keep up. The World Health Organization has issued guidelines for Large Language Models (LLMs); the European Union has implemented the AI Act, a first-of-its-kind attempt to comprehensively regulate AI and ensure that its development and deployment are responsible, trustworthy, and secure; and the FDA is advancing frameworks for the lifecycle management of AI-based medical devices. In Israel, too, the government approved a national AI program in 2023, now in its second phase, together with investments in healthcare AI.
The wheels of regulation move slowly, and meanwhile technology is already making changes on the ground. “AI in medicine must be used wisely and under strict oversight,” warns Professor Ido Wolf, Head of the Oncology Division at the Tel Aviv Medical Center, and Head of Tel Aviv University Medical School. “What’s to stop a resident from inputting patient data into an AI chat in the middle of the night, and following its recommendations? Or prevent patients from entering everything into Chat themselves, and arriving to their doctor with a detailed set of demands?”
Smart Clinical Documentation – Less Bureaucracy, Less Burnout, Greater Efficiency
GenAI systems are already changing clinical workflows. Doctors use automated tools to summarize visits, generate discharge letters, and translate patients’ language into professional medical terminology. The result is less red tape, reduced staff burnout, and faster, more efficient care.
One example of how AI is supporting the delivery of healthcare is a system developed by Matrix DnA for a major HMO. “The system we developed uses GenAI to analyze hospital visit summaries that are received from hospitals, and automatically check documents,” explains Asaf Timor, CEO of Matrix DnA. “In many countries, when a patient goes to hospital they must bring with them proof of their HMO’s approval to pay for treatment. From their perspective, once they hand this in, the issue of payment is closed, but for the HMO it’s just begun. Without a well-organized interface with the hospitals, medical review teams, including doctors and interns, are required to go through every set of documents submitted, and compare the medical information sent by the hospital with the payment request sent by the hospital. This manual process takes up valuable time for doctors and wears out the medical team. The solution that we’ve developed knows how to identify whether the bill is correct or not, and provides a clear and reasoned explanation which enables quick handling of errors and an expedited appeal process, if necessary. It also recovers funds that would otherwise fall through the cracks, saving the HMO millions of shekels a year.”
Clinical and Epidemiological Management, Systems Operation, and Smart Hospitals
According to Bruno Lavi, one of the fields in which AI has already proven itself is clinical and epidemiological management in healthcare and hospital systems. “Algorithms enable optimization of surgical and appointment schedules, bed allocation, resource prioritization, and staff management, according to real-time needs. For example, in the US, AI systems are being used to optimally manage emergency rooms, shortening wait times by tens of percent. Some NHS hospitals in England have even begun testing AI-based solutions to continuously analyze real-time patient data to predict seasonal demand, so that staff and equipment can be prepared in advance.”
“Another area in which the AI revolution is leading significant change is the issue of hospital operations, including the fields of engineering and logistics,” says Bruno. “Hospitals face enormous challenges every day that are not always directly related to the allocation of resources and tools required for operational management. In recent years, capabilities and systems have begun to appear that can more effectively handle operational processes, perform real-time control, manage smart equipment, handle cleaning, and improve energy efficiency. The world of AI has even begun to directly impact special operational issues, such as violence against staff or even a terrorist attack.
“Video analytics technology enables real-time analysis of video content, identifying patterns and anomalies, and generating actionable insights. While it was previously used primarily in clinical contexts, such as monitoring body movements in rehabilitation, or detecting patient falls, it now also plays a central role in hospital operations management, helping monitor emergency room overload, improve staff safety through real-time risk identification, and control the efficient utilization of resources, such as operating rooms and medical equipment.”
AI in a Gown: Diagnosis and Treatment
The revolution doesn’t stop at logistics. According to research carried out in Israel, ChatGPT can pass medical exams and outperform many residents in internal medicine and psychiatry. Another study, conducted at an emergency medical center in the US, found that, in common conditions such as respiratory or urinary tract infections, AI was able to diagnose and offer treatment recommendations more accurately than doctors. This means that artificial intelligence is already capable of performing some of the doctor’s work with greater accuracy and speed, and in fact is dramatically changing the medical profession.
“We recently developed an early cancer detection tool for startup Genesis Medical,” says Or Lustig, VP of Business Development at Matrix Medica. “Our AI platform learns and imitates the logic and intuition of an expert doctor, thus identifying cancer in the earliest stages with unprecedented levels of reliability and accuracy. The company’s first product, G4Lungs, is designed to screen for lung cancer, the deadliest and second most common cancer. In a clinical trial, it showed extraordinary results: 97 percent accuracy.
“Thanks to this success, the product has entered a commercial pilot that is fully integrated into the PACS system, as part of a strategic project with the world’s leading organization in the field of lung cancer, across 11 European countries. The company is also developing a product for the early detection of prostate cancer, which aims to significantly reduce unnecessary biopsies, as part of the vision to establish a broad platform for the early diagnosis of all types of cancer.”
Preventive and Personalized Medicine
For patients, one of the most exciting things about AI in healthcare is preventive and personalized medicine. In Israel, HMOs are utilizing historical clinical databases on an unprecedented scale to identify patients who are at risk, even before symptoms appear. For example, Maccabi’s KSM Institute published innovative studies that demonstrated the ability to identify the development of celiac disease years in advance, using machine learning models.
The preventive medicine approach is also gaining momentum worldwide. In the US, Kaiser Permanente uses algorithms to identify early risk of cardiovascular events and suicide, by cross-referencing laboratory data, psychiatric records, and health habits. In the UK, the NHS has implemented AI-based models to predict complications during childbirth, and for early identification of patients at risk of developing dementia.
What these initiatives have in common is the combination of advanced predictive capabilities and the personalization of treatment, so that medicine not only responds to the disease, but also anticipates it, changing according to the patient’s personal profile – a revolution that can save lives, conserve resources, and improve the quality of life of millions of people.
Challenges and Barriers: Balancing Innovation and Responsibility
The implementation of AI systems in the healthcare system comes with a range of fundamental regulatory, clinical and ethical challenges. A main one is the quality of the data and the biases inherent in them. Medical databases often reflect historical, social or gender bias, which may lead to models replicating and even exacerbating existing gaps.
Another challenge concerns explainability and clinical trust; doctors need to understand when and how to rely on the recommendation of an AI system, especially when it comes to life-threatening decisions. The guidance of the FDA and the European Union emphasize the importance of including a human in the decision-making process, and the need for ongoing monitoring after implementation.
In addition, issues of data privacy and secondary use of data continue to cause public concern. For example, in the UK, unregulated chatbots were reported to be operating in health and welfare services, using sensitive data, without sufficient supervision.
Where the Healthcare Revolution Goes from Here
Generative AI is not just upgrading the healthcare system; it is changing the entire paradigm. We are moving from retrospective insights based on historical data, to an era of real-time assistance for doctors, operational teams and, most importantly, patients.
Israel, which combines decades of medical information infrastructure, leading innovation institutions, and an entrepreneurial spirit, is uniquely positioned to become a global arena for the responsible and groundbreaking implementation of GenAI in healthcare.