According to a report by SNS Insider published in March 2026, the global AI in healthcare market was valued at 38.01 billion USD in 2025, and is projected to reach an astonishing 1,222.12 billion USD by 2035, with a compound annual growth rate (CAGR) of 41.50%. This is not just a statistic — it is a signal that the entire medical industry is undergoing a fundamental transformation.

The medical diagnostics segment is particularly significant. According to Precedence Research, the global AI in diagnostics market is worth 2.17 billion USD in 2026 and is growing toward 15.74 billion by 2035. Key players in this market include companies such as Siemens Healthineers, GE Healthcare, Aidoc, Google Health and Paige.

Radiology is the field where AI has made the greatest breakthrough. As many as 75–80% of all FDA-approved AI medical devices are in the field of radiology, according to data from December 2025. The reason is clear: medical image analysis is a perfect task for machine learning — large volumes of data, recognizable patterns, and a need for speed and precision.

In 2024, more than half of healthcare providers were actively using AI for at least one medical imaging task, compared to just 17% in 2018. A 2024 European radiologist survey found that 48% of respondents actively use AI tools, with another 25% planning to do so.

Breast cancer is the most common cancer in the world — according to WHO estimates, it accounts for 12.5% of all new annual cancer cases, and claims more than 670,000 lives every year. Early diagnosis is critical: if diagnosed at an early stage, more than 90% of patients can be cured, and the five-year survival rate is 96%.

This is precisely where AI demonstrates its true value. A study by researchers at the University of Aberdeen showed that an AI tool increases the breast cancer detection rate, particularly for forms that are harder to spot on traditional screenings. Early results show that AI can be especially useful as a supplement to medical staff, not a replacement — the system works in collaboration with radiologists, shortens analysis time and reduces the number of missed cases without increasing false positive diagnoses.

At the regional level, the JMM team from the AI4Health.Cro innovation competition developed a prototype application that uses artificial intelligence to identify potentially cancerous lesions on breast mammograms according to the BI-RADS system, which is important for assessing the degree of breast cancer risk. This shows that the region is beginning to actively work on AI solutions for healthcare.

Google DeepMind and Google Research have developed a family of medical AI models that represent a new frontier in the application of artificial intelligence in healthcare. Med-Gemini is a clinical AI model that supports medical diagnostics, interpretation of radiological images, treatment planning, summarization of electronic health records (EHR) and analysis of genomic data.

The concrete results are impressive: MedGemma 4B achieves 64.4% on the MedQA benchmark, placing it among the best models below 8 billion parameters. In a non-blinded study, 81% of chest X-ray reports generated by MedGemma 4B were rated by US board-certified radiologists as sufficiently accurate to make similar patient management decisions as the original radiologists. On the MedQA (USMLE-style) benchmark, Med-Gemini achieves an accuracy of 91.1%, setting a new industry benchmark.

MedGemma 1.5 4B also brings concrete improvements in anatomical localization — a 35% improvement in intersection over union on the Chest ImaGenome benchmark. All models can be run on a single GPU, and MedGemma 4B can be adapted even for mobile hardware.

Personalized medicine — or precision medicine — is an approach in which treatment is tailored to the individual characteristics of each patient: their genetic profile, microbiome, lifestyle and medical history. AI is the key driver of this revolution.

AI cancer detection tools have achieved a 93% match rate with expert tumor board recommendations, helping healthcare workers make decisions based on the unique characteristics of each patient. The AI in precision medicine market is projected to reach 14.5 billion USD by 2030.

Med-Gemini processes data from various sources — electronic health records, medical images, laboratory results and patient-generated data — to predict disease progression, potential complications and treatment outcomes. AI analysis of hundreds of exomes in medulloblastoma cases identified specific molecular subgroups, enabling physicians to apply precise treatment doses.

Siemens Healthineers presented a new suite of AI-powered radiology services at the RSNA 2025 conference in Chicago, covering the entire imaging chain — from scheduling to report generation. Pilot projects showed that radiologists can annotate chest CT scans up to 25% faster using the AI solution, and cognitive load was reduced by at least 16%. The company generated revenues of approximately 23.4 billion euros in fiscal 2025 with around 74,000 employees worldwide.

Aidoc, an Israeli clinical decision support startup, has raised 370 million USD and signed more than 150 contracts with health systems such as Mount Sinai, Yale New Haven Health and Sutter Health. Aidoc's intracranial hemorrhage (ICH) detection tool achieves more than 90% sensitivity with low false positive rates. The aiOS™ platform integrates deeply with EHR, PACS, scheduling and reporting systems.

According to data from mid-2025, the FDA has approved a total of approximately 873 algorithms for radiology AI, with leading vendors including GE Healthcare (96 approved tools), Siemens Healthineers (80), Philips (42) and Aidoc (30).

The regulatory landscape for AI in healthcare is undergoing a major transformation. The EU AI Act entered into force in August 2025, and from August 2026 most provisions for high-risk AI systems become binding. Medical AI devices that serve as safety components or are themselves regulated products under the EU MDR are classified as high-risk AI systems, requiring technical documentation, risk management, human oversight and transparency.

Penalties for violating the EU AI Act can reach up to 7% of a company's annual revenue. For healthcare organizations and manufacturers of medical AI devices, this means that proactive compliance is not only an ethical obligation but also a strategic necessity. Montenegro, as an EU candidate country, will face increasing pressure to align its regulations with these standards.

Montenegro is actively positioning itself in the global AI economy. At the Second National Conference "AI DIRECTLY: Montenegro in the Age of AI Challenges" in December 2025, Prime Minister Milojko Spajić presented the Government's strategic vision focused on accelerated digital development and positioning Montenegro in the global AI economy, including the application of AI in healthcare. A panel discussion on the application of AI in healthcare was one of the three central thematic blocks of the conference.

UNDP's AI Landscape Assessment (AILA) report for Montenegro from 2024 shows that the country has a well-established regulatory framework for data governance (score of 4.6 out of 5 — transformational phase), providing a solid foundation for the responsible introduction of AI in healthcare. However, digital infrastructure and AI-specific capacities still require investment.

The European Commission has launched the SHAIPED project, which since March 2025 has been piloting the development, validation and deployment of AI models using the HealthData@EU infrastructure of the European Health Data Space (EHDS) — which entered into force in 2025. Montenegro, as part of its EU integration path, will have the opportunity to connect with these infrastructures and access advanced AI solutions for diagnostics.