Original article (in Croatian) was published on 18/9/2025; Author: Ivan Nekić
The Ministry of Health points out that artificial intelligence (AI) is already present in the Croatian health system, and they also gave us examples of the use of new technology.
Artificial intelligence (AI) is slowly entering all spheres of our lives. Today, we can hardly imagine life without our virtual assistants powered by artificial intelligence – ChatGPT, Perplexity, Grok, Deepseek and many others. We can ask these assistants for a recipe suggestion based on what we have in the fridge, as well as for advice for a potted plant wilting on your windowsill or terrace. Artificial intelligence has entered our lives so deeply that there are now refrigerators that use a camera to record the interior and display the expiration dates and number of groceries on the external screen, and they also send reminders if you need to buy something in the store.
AI is also increasingly entering the health care system, i.e. hospitals, health centres and outpatient clinics. For example, the first virtual AI hospital was opened in China in 2024. In the virtual world, virtual patients are treated with artificial intelligence. The AI hospital aims to train doctors through a simulated environment in order to be able to autonomously develop and improve the ability to treat the disease. The AI hospital called Agent Hospital was developed by researchers from Tsinghua University.
Researches point out that AI doctors can treat 10,000 patients in two days, which would take doctors in the real world about two years. Doctors powered by artificial intelligence thus simulate the entire process of diagnosis and treatment of patients, including consultation, examination, diagnosis, treatment and control.
The question is whether real doctors can have any benefits from developing an AI hospital with virtual doctors? The researchers who developed this model believe they can. The head of research on the Agent Hospital project, Liu Yang, says the AI hospital city should transform the way doctors diagnose and treat patients, bringing great benefits to both medical professionals and the general public.
Ministry of Health: There is a significant presence of AI in our healthcare system
Given the very rapid development of AI technologies in all sectors, we were interested in the degree to which artificial intelligence is present in Croatian healthcare. We asked the Ministry of Health how they see the role of artificial intelligence in Croatian healthcare system in the next 5 to 10 years. The Ministry points out that artificial intelligence is already significantly present in the Croatian health system, especially through software solutions integrated into medical devices in public hospitals.
“Examples include robotic surgical systems, applications for radiological imaging analysis, and digital pathology systems. We expect further growth in AI application, supported by the European Digital Innovation Centre AI4Health.Cro and the implementation of the recently adopted Regulation establishing a European Health Data Space. Consequently, in the coming period, we certainly expect national regulations to be harmonised with this Regulation, in order to ensure the privacy and safety of patients within data processing, and on the other to enable the development of innovative solutions that further improve the efficiency of health care, while always preserving the safety of patients,” the Ministry pointed out.
In the Croatian health system, AI is used as an auxiliary tool in various areas, always under the supervision of professional staff and further highlighting several examples. “In robotic surgery, systems for urological, abdominal, gynaecological and thoracic surgery are applied, which enable advanced image analysis and support in performing the procedure. For example, a robotic system delivered in 2019 for urological surgery, especially in prostate cancer, is used in one clinical hospital centre, while in 2024 a new system used by urology and abdominal surgery teams was procured, with a plans to include head and neck surgery, gynaecology and thoracic surgery. The systems record and analyse the doctor’s performance indicators, helping to improve the effectiveness and outcomes of the procedure,” the Ministry explained.
The added that applications for early screening of lung cancer with AI support in preanalysis of images is used in radiology, with decisions always made by the doctor, and data processed locally. In the field of radiotherapy, with the implementation of new linear accelerators and a centralized oncology information system, specialized AI-powered tools are now available to suggest and mark organ contours, enabling precise therapy planning and application under the constant supervision of a doctor.
Similarly, digital scanners of pathological samples supported by AI in the analysis of breast cancer are used in pathology, with each result undergoing additional review by a pathologist before the final report,” the Ministry points out. It emphasizes that in all these cases, artificial intelligence serves only as an aid to medical professionals, while doctors retain full responsibility for diagnosis and treatment—ensuring that technology enhances the quality of care without replacing the profession.
AI4HealthCro – consortium for the development of AI-based healthcare solutions
We asked one of the partners in the consortium – the Croatian Institute of Public Health (HZJZ) – to explain what is AI4Health.Cro. “It is a non-profit, public-private consortium formed in 2023 that sees artificial intelligence (AI) as key to the advancement of healthcare and is working on developing innovative digital solutions that use artificial intelligence or AI solutions that can help doctors,” they explained.
The consortium, according to HZJZ, brings together 15 partners under the coordination by the Ruđer Bošković Institute (IRB), and the Croatian Institute of Public Health is one of the partners. The AI4Health.Cro centre, among other things, encourages startups through innovation competitions, and supports numerous other companies in their digital transformation process, thanks to cooperation with a consortium of 16 sector partners. In April this year, at the annual conference of the European Digital Innovation Hub for the Application of Artificial Intelligence in Health and Medicine (EDIH AI4Health.Cro), AI4Health.Cro project coordinator Anja Barešić pointed out that the AI4Health.Cro centre not only encourages startups through innovation competitions, but also provides key support to a number of other companies in their digital transformation process, thanks to cooperation with a consortium of 16 sector partners. “We are satisfied with the conclusion of the project’s second year because our services were used by 37 users, each with a personalized digital transformation program that includes education, networking, financing and testing,” Barešić pointed out.
To encourage innovation in healthcare through the use of artificial intelligence, the AI4Health.Cro consortium announced an innovation competition in February 2024, offering applicants the chance to win prizes from a fund of EUR 9,000. The aim of the competition was to develop an innovative solution to the problem of early rehospitalization using artificial intelligence. Specifically, innovators were asked to use artificial intelligence to develop a solution that would provide for the hospitalization of a patient up to 30 days after discharge from the hospital.
The winning team at the competition consisted of alumni of the Faculty of Electrical Engineering and Computing (FER) and the Faculty of Science of the University of Zagreb, and the task was to use AI for the analysis of mammographic images, i.e. early detection of breast cancer. Early diagnosis is key to saving lives, and timely mammography exams through programs for early detection of breast cancer have proven necessary.
Competitors first needed to develop an AI model that recognizes and classifies breast lesions according to the BI-RADS system, which is important for assessing the degree of risk of breast cancer. Then they had to mark the suspicious areas on the scans, just like an experienced radiologist does. In the third task, they had to explain how their model arrives at conclusions, which helps to detect important patterns and possible biases. Finally, the fourth task was to design a simple and useful interface that makes it easier for doctors to interpret the results, with clear explanations of the diagnoses and their reliability.
Is a greater implementation of AI in Croatian health care planned?
As of 2 August 2026, the European Union’s AI Act will be in full force and effect, making it the world’s first comprehensive legal framework for AI regulation. As it is pointed out, its goal is to simultaneously encourage innovation and ensure the safe, ethical and reliable use of AI systems in the EU. It will also introduce special rules for generative AI (ChatGPT, Perplexity and other platforms of this type). We were interested in how the Ministry of Health is preparing for the implementation of this act. They briefly replied that some provisions of the act are in force already and that the harmonization of national regulations with EU regulations will ensure the safe development of innovations in health care, with full protection of patients and their data and the responsibility of authorized health professionals.
Although not much can be surmised from their answer about their plans, they nevertheless point out that, in accordance with the development of technology, a greater implementation of systems powered by artificial intelligence is expected.
“As AI models and algorithms mature for specific application areas, procedures that ensure their reliability and security will evolve. Machine and deep learning approaches will be implemented, maintained and upgraded to ensure the highest possible quality and safety for patients. Priorities in health and AI-based solutions will be highlighted in the new national strategy and the plan for the use of artificial intelligence, which is being prepared by a comprehensive interdepartmental working group,” the Ministry explained.
Although artificial intelligence is not explicitly mentioned in the Implementation Programme for 2024-2028, they said that the development of digital tools and the digitization of health infrastructure are certainly implied. They pointed out that AI is considered a tool that, with proper certification and supervision, can improve certain segments of health care, but cannot be a comprehensive solution to all challenges in the health system.
When asked about the readiness of health data in Croatia for use in AI projects, the Croatian Institute of Public Health highlighted that the Regulation on the European Health Data Space (EHDS), which entered into force on 26 March this year, represents a key step toward harmonizing access to and sharing of health data within the European Union. “This regulation further underscores the importance of developing secure and interoperable data processing systems, to which the HZJZ actively contributes through its projects,” the Institute stated.
Hospitals and institutes in Croatia participating in projects bringing AI into healthcare
Artificial intelligence, in addition to clinical and diagnostic medicine, can also help in the field of epidemiology and preventive medicine. In the field of preventive medicine, digital AI assistants can help educate, inform and increase the health literacy of patients and citizens. Innovations using AI are already present worldwide in many areas of medicine and health, for example in diagnostics, clinical care for patients, drug development, epidemiological monitoring and disease control, interventions to combat and prevent epidemics, as well as the organization and management of health systems, the Institute points out, citing similar projects introducing AI in Croatian health care:
“For example, colleagues at Dubrava Clinical Hospital are working on the development of a digital assistant to help inform patients about guidelines and recommendations for the treatment of cardiac diseases according to their needs and health condition. Also, the Croatian Institute of Public Health is a partner in the development of the digital assistant for gynaecological literacy, Gina-AI, which was made in cooperation with the AI marketing agency Woom, the communication company Infobip and numerous health professionals. It is an innovative digital solution that uses artificial intelligence to provide general, verified and understandable information in the field of gynaecology, including some topics in sexual and reproductive and mental health for educational and informational purposes. “
They also note that they are currently involved in the EU Horizon project application in the field of AI and pandemic response preparedness (HORIZON: Leveraging Artificial Intelligence for Pandemic Preparedness and Response), as one of the partners, along with the Clinic for Infectious Diseases “Dr. Fran Mihaljević”, Infobip and Statens Serum Institute.
How is AI used in medicine in high-income countries around the world?
In order to try to compare the implementation of AI in Croatian health care compared to high-income countries around the world, we briefly explored the type of projects based on artificial intelligence implemented in these countries.
In the UK, for example, mammography screening programs for breast cancer use artificial intelligence systems. A NHS (UK national health service) AI tool called Mia detected approximately 12% more cases of cancer than used to be the practice, reduced the number of erroneous invitations for additional examinations, and is projected to be able to reduce the radiologists’ workload by up to approximately 30%. The evaluation of this tool is ongoing, and a large national survey is planned to include almost 700,000 women in the UK to more reliably explore how useful AI tools can be to detect breast cancer earlier and consequently reduce mortality from this cancer.
Also, in some hospitals in the UK, a project in which AI participates in the analysis of tissue samples – Galen Breast – is in the trial phase. The project uses algorithms to analyse breast biopsies for faster and more accurate diagnosis and to prioritize more urgent cases. The project will actually help show if Galen Breast can make breast cancer assessment faster and more accurate, leading to earlier treatment and freeing up time for clinicians.
In Japan, robotic systems powered by artificial intelligence are used to help care for the elderly. The AIREC robot, developed at Waseda University, is used as a humanoid assistant for repositioning patients (e.g. dressing, preventing pressure ulcers), changing clothes, and support more demanding physical tasks in care settings when the workforce is limited. However, it should be noted that this is only a prototype robot that is in the research and testing phase under controlled laboratory conditions. However, the results so far are very promising.
German AI platform Elea uses AI to analyse medical images, reducing the time needed for diagnosis from weeks to several hours. This approach enables faster decision-making and better patient care. A French company (Implicity) uses AI to remotely monitor patients with cardiovascular disease. The system allows monitoring of nearly 80,000 patients in the US and Europe, reducing the death rate by 26%.
According to the World Economic Forum, 4.5 billion people in the world do not have access to basic health services, and AI can significantly help bridge that gap. Already today, AI is helping doctors detect fractures, triage patients, and diagnose diseases early. With the expected shortage of workforce in the healthcare system, the role of this technology is becoming increasingly important.
Nevertheless, AI should remain an assistant, while doctors and medical staff should continue to be in charge. Their knowledge and experience in treatment, surgery and prevention can hardly be replaced by technology. With the introduction of AI in the health system, Croatia does not lag significantly behind most European countries, but a comparison with high-income countries such as China, Japan and the USA shows that these projects already have a strong impact on the way care is provided there.
