Ukrainian Healthcare System: Current Medical Imaging Practices

---

Abstract #

Ukraine’s healthcare system represents a unique case study in digital transformation under extraordinary circumstances. The two-level electronic healthcare system (EHS), with 36 million registered patients and 1.6 billion electronic medical records, provides a robust foundation for AI integration—despite wartime challenges that reduced viable Medical Information System (MIS) providers from 40 to just 4-6. This article examines Ukraine’s current medical imaging infrastructure, analyzes the digital health architecture enabling AI deployment, and identifies strategic opportunities for ML-based diagnostic systems. The findings reveal that while equipment density falls below EU standards, the consolidated MIS landscape and standardized API infrastructure create favorable conditions for systematic AI implementation.

---

1. Historical Context: From Soviet Legacy to Digital Transformation #

Ukraine inherited a declining Soviet healthcare system after independence in 1991. Despite constitutional promises of universal, free healthcare, the system remained chronically underfunded, bureaucratically inefficient, and dependent on informal “care for cash” arrangements. Political connections often determined access to quality care rather than clinical need, creating significant healthcare inequities across the population.

The first major reform wave began in 2014, accelerated by the Euromaidan revolution and subsequent political realignment toward European integration. In 2017, the National Health Service of Ukraine (NHSU) was established as a single-payer system, fundamentally restructuring healthcare financing. Critically, the electronic healthcare system (EHS) was mandated as the sole platform for all medical service providers, creating a unified digital infrastructure that would prove essential for future AI integration.

The healthcare reform process continued through 2020-2024, with phased implementation of secondary and tertiary care reimbursement through the NHSU. By late 2025, the eZdorovya platform had processed nearly 5 billion medical data records, earning the GovTech Award 2025 and demonstrating Ukraine’s capacity for rapid digital health infrastructure development even amid conflict conditions.

timeline
    title Ukrainian Healthcare Digital Transformation
    1991 : Independence
Soviet system inherited
    2014 : Reform begins
Post-Euromaidan changes
    2017 : NHSU established
Single-payer system
    2019 : EHS mandatory
Digital platform launch
    2022 : War begins
MIS consolidation
    2025 : 5B records
GovTech Award

Figure 1: Timeline of Ukrainian Healthcare Digital Transformation

2. Ukraine’s Two-Level Digital Healthcare Architecture #

The Ukrainian EHS employs a distinctive two-level architecture that balances centralized data governance with local implementation flexibility. This design provides significant advantages for AI integration, as it establishes unified data standards while allowing regional customization for specific clinical workflows.

flowchart TD
    subgraph Central["CENTRAL DATABASE"]
        A["36M Patients"]
        B["1.6B Records"]
        C["400K Users"]
        D["1000-1500 req/sec"]
    end
    
    Central --> API["Standardized API"]
    
    API --> MIS1["MIS Provider 1"]
    API --> MIS2["MIS Provider 2"]
    API --> MIS3["MIS Provider N"]
    
    MIS1 --> H1["Hospital A"]
    MIS2 --> H2["Clinic B"]
    MIS3 --> H3["Imaging Center C"]
    
    style Central fill:#e3f2fd,stroke:#1976d2
    style API fill:#fff3e0,stroke:#f57c00

Figure 2: Ukraine’s Two-Level Electronic Healthcare System Architecture

Central Component Features:

• National Data Repository: Unified storage for patient records, prescriptions, and clinical outcomes
• Unified Registries: Standardized databases for patients, healthcare facilities, and health professionals
• Processing Capacity: 1,000-1,500 API requests per second, supporting real-time clinical workflows
• Interoperability Standards: All MIS providers must transfer data in identical formats to the central database

Local MIS Layer Features:

• Pre-War Diversity: Over 40 medical information systems developed by various IT companies
• Regional Customization: Local systems can create unique functions for specific regional healthcare needs
• Clinical Workflow Integration: Direct integration with hospital EHR systems, laboratory information systems, and PACS
• Post-War Consolidation: Reduced to 4-6 viable providers, actually simplifying integration requirements

3. Medical Imaging Equipment: The Infrastructure Gap #

Ukraine’s medical imaging infrastructure reveals a significant equipment density gap compared to European Union standards. This disparity affects both the availability of diagnostic imaging services and the volume of training data available for AI model development.

xychart-beta
    title "CT Scanners per 100,000 Population (2022)"
    x-axis ["Greece", "Germany", "EU Avg", "Hungary", "Ukraine"]
    y-axis "Units per 100K" 0 --> 6
    bar [4.9, 3.5, 2.8, 1.1, 1.2]

Figure 3: CT Scanner Density Comparison (Source: Eurostat 2022, Ukrainian estimates)

The equipment comparison reveals several concerning patterns:

Country	CT Scanners per 100K	MRI Units per 100K
Greece (highest EU)	4.9	3.7
Germany	3.5+	3.0+
EU Average	~2.8	~1.8
Hungary (lowest EU)	1.1	<1.0
Ukraine (estimated)	~1.0-1.5	~0.5-0.8

War Impact on Infrastructure: Since February 2022, conservative estimates indicate 707 documented assaults on healthcare facilities and 86 direct attacks on medical personnel. Many imaging centers in eastern and southern regions have been destroyed or forced to shutter operations, further reducing already limited capacity. The concentration of surviving advanced imaging equipment in western Ukraine (Lviv, Ivano-Frankivsk) and the capital (Kyiv) creates significant geographic access disparities.

4. Telemedicine Acceleration During Conflict #

The ongoing conflict has paradoxically accelerated telemedicine adoption across Ukraine, creating new infrastructure that supports remote AI-assisted diagnostics. Conflict-affected oblasts demonstrate the highest telemedicine utilization rates, driven by necessity rather than planned digital health strategy.

Regional Adoption Patterns:

• High Adoption (conflict-affected): Kyiv, Chernihiv, Kharkiv, Sumy, and Kherson oblasts show the highest telemedicine utilization
• Pre-War Infrastructure Leaders: Odesa, Lviv, and Poltava oblasts had established telemedicine centers before 2022
• Provider Distribution: Private healthcare facilities provide the majority of telemedicine services, with public sector lagging

Service Modalities in Use:

• Synchronous video consultations (most common)
• Asynchronous medical data sharing for specialist review
• Audio-only consultations for low-bandwidth regions
• Text-based consultations and prescription management

The Diia mobile application, launched in July 2024, now enables patient access to personal eHealth records, creating a direct consumer interface to the national health data infrastructure. Over 31.6 million Ukrainians have signed declarations with primary care physicians through the system, representing approximately 85% of the population.

5. Digital Health Infrastructure for AI Integration #

The existing EHS registries provide multiple AI-ready data sources that can support various machine learning applications in healthcare:

Registry	Data Type	AI Application Potential
Patient Registry	Demographics, declarations	Population health analytics, risk stratification
Healthcare Facilities	Locations, capabilities, equipment	Resource optimization, capacity planning
Health Professionals	Specializations, workload data	Workflow analysis, staffing optimization
Prescriptions	Medications, dosages, histories	Drug interaction detection, adherence prediction
Electronic Medical Records	Clinical notes, diagnoses, outcomes	Diagnostic AI training, clinical decision support
COVID-19 Data	Testing, vaccination, outcomes	Epidemiological modeling, outbreak prediction

Upcoming Digital Health Initiatives (2025-2027):

• e-Stock: Digital inventory management for medical supplies enabling supply chain optimization
• State Registry of Medicines: Modernization project for comprehensive drug tracking
• Medical Equipment Registry: Device tracking including imaging equipment maintenance and utilization
• Rehabilitation Initiative: Specialized tracking for wounded civilian and military care
• Big Data Analysis Platform: Dedicated infrastructure for ML applications in healthcare

6. Critical Challenges for AI-Powered Medical Imaging #

Despite the promising digital infrastructure, several critical constraints affect AI implementation in Ukrainian medical imaging:

mindmap
  root((AI Imaging
Challenges))
    Infrastructure
      Equipment scarcity
      Facility destruction
      Geographic concentration
    Workforce
      IT specialist emigration
      Radiologist shortages
      Training gaps
    Technical
      MIS consolidation
      Data quality variance
      Interoperability issues
    Financial
      IT company losses
      Limited investment
      Currency instability

Figure 4: Critical Challenges for AI-Powered Medical Imaging in Ukraine

Key Constraint Analysis:

1. MIS Provider Consolidation: Pre-war ecosystem of 40+ MIS providers reduced to 4-6 viable systems, requiring focused integration efforts but also simplifying standardization
2. IT Workforce Displacement: Many qualified software specialists and data engineers fled abroad during 2022-2023, creating talent gaps in health IT
3. Infrastructure Destruction: Imaging centers in conflict zones damaged or destroyed, reducing both service capacity and data generation
4. Financial Constraints: IT companies serving the healthcare sector suffered significant losses, limiting investment in AI development
5. Equipment Scarcity: CT/MRI density already below EU standards, further reduced by war damage

7. Strategic Opportunities for AI Integration #

Despite challenges, Ukraine’s unique circumstances create several strategic opportunities for AI-powered medical imaging deployment:

The “Leapfrog Opportunity”: Ukraine’s forced digital transformation during crisis could enable faster AI adoption than gradual EU approaches—similar to mobile banking leapfrogging in developing economies. Countries that missed legacy infrastructure investments sometimes adopt newer technologies faster than incumbents.

MIS Consolidation Advantage: While tragic in its origins, the reduction from 40 to 4-6 MIS providers actually simplifies AI integration. Fewer systems to support means more standardization, reduced testing burden, and clearer partnership pathways.

Private-Public Imbalance as Entry Point: The private sector’s dominance in telemedicine suggests AI solutions should target private facilities first, then expand to public hospitals. Private facilities have more flexibility in technology adoption and faster procurement cycles.

Data Richness Despite Chaos: The combination of 1.6 billion electronic medical records, comprehensive COVID-19 tracking data, and emerging rehabilitation datasets creates substantial training data for Ukrainian-specific ML models. This data reflects real-world conditions including conflict-related trauma patterns.

Workforce-AI Synergy: With medical workforce depleted by emigration, casualties, and displacement, AI assistance transitions from luxury enhancement to operational necessity for maintaining care quality. This creates stronger adoption incentives than in well-staffed healthcare systems.

8. Regional Healthcare Infrastructure Analysis #

Understanding the regional distribution of healthcare resources is essential for strategic AI deployment planning. Ukraine’s 24 oblasts plus Kyiv city exhibit significant variation in healthcare infrastructure, equipment availability, and digital health adoption rates.

Western Ukraine: The Stability Corridor #

The western oblasts—Lviv, Ivano-Frankivsk, Zakarpattia, Chernivtsi, Ternopil, Khmelnytskyi, and Volyn—have emerged as Ukraine’s healthcare stability corridor. These regions experienced minimal direct conflict impact and have absorbed significant internal migration, increasing healthcare demand while maintaining infrastructure integrity.

Lviv Oblast has emerged as a particular focal point for healthcare development, hosting multiple international humanitarian medical programs and receiving equipment donations from European partners. The region’s medical universities continue operating at full capacity, maintaining the physician training pipeline critical for long-term workforce sustainability. Several private imaging centers in Lviv have invested in modern equipment, creating potential pilot sites for AI-assisted diagnostics.

Ivano-Frankivsk and Zakarpattia oblasts serve as critical nodes for cross-border medical evacuation routes to Poland, Hungary, and Romania. This positioning has driven investment in emergency care infrastructure and established international medical cooperation frameworks that could support AI technology transfer initiatives.

Central Ukraine: The Reconstruction Zone #

Central oblasts including Kyiv, Zhytomyr, Cherkasy, Vinnytsia, and Poltava represent a mixed landscape of damaged and operational healthcare facilities. Kyiv Oblast experienced significant early-war damage but has seen substantial reconstruction investment. The capital city of Kyiv maintains Ukraine’s highest concentration of specialist physicians and advanced diagnostic equipment, though many facilities required repairs following 2022 attacks.

Poltava Oblast represents a particularly interesting case study for AI deployment. The region maintained relatively stable healthcare operations throughout the conflict and has an established telemedicine center predating 2022. The oblast’s healthcare administration has demonstrated openness to digital health innovations, making it a potential testbed for AI diagnostic pilots.

Southern and Eastern Ukraine: Conflict Impact Zones #

The southern oblasts (Odesa, Mykolaiv, Zaporizhzhia, Kherson) and eastern oblasts (Kharkiv, Dnipropetrovsk, Donetsk, Luhansk) have experienced the most severe healthcare infrastructure impacts. While some areas remain under Russian occupation with unknown healthcare conditions, liberated territories reveal systematic destruction of medical facilities.

Odesa Oblast presents a unique opportunity despite its proximity to conflict zones. The city of Odesa maintains functioning healthcare infrastructure, hosts an established telemedicine center, and serves as a hub for medical evacuations from southern regions. The oblast’s academic medical institutions continue research programs and maintain connections with international partners, creating favorable conditions for AI technology introduction.

Kharkiv Oblast has demonstrated remarkable healthcare system resilience despite intensive bombardment. Medical facilities have adapted to operating under constant threat, developing rapid patient evacuation protocols and distributed care delivery models. These adaptations, while born of necessity, have created healthcare delivery innovations that could inform AI-assisted triage and resource allocation systems.

9. Workforce Considerations for AI Implementation #

The successful deployment of AI-powered medical imaging depends critically on workforce readiness—both the technical staff to implement and maintain systems and the clinical staff to effectively utilize AI assistance in diagnostic workflows.

Radiologist Workforce Status #

Ukraine’s radiologist workforce faced significant disruption during 2022-2024. Pre-war estimates suggested approximately 4,500-5,000 practicing radiologists across the country, already below European per-capita standards. The combination of emigration, conscription, and displacement has reduced this number by an estimated 15-25%, with the precise figure difficult to verify given ongoing conflict conditions.

The workforce reduction creates a paradoxical opportunity for AI adoption. In well-staffed healthcare systems, AI diagnostic assistance may face resistance from radiologists concerned about workload impact or professional displacement. In Ukraine’s current context, AI systems that can amplify radiologist productivity directly address an acute staffing crisis, creating stronger adoption incentives and reduced implementation resistance.

Health IT Workforce Challenges #

Ukraine’s pre-war IT sector was among the most developed in Eastern Europe, with an estimated 200,000+ software developers and a thriving outsourcing industry. The healthcare IT sector, while smaller, benefited from this broader ecosystem. However, the IT sector experienced significant workforce displacement during 2022, with estimates suggesting 30-50% of IT professionals relocated abroad at least temporarily.

This displacement particularly affected health IT companies serving the MIS market. The reduction from 40+ to 4-6 viable MIS providers reflects both company closures and workforce losses. The surviving MIS providers have consolidated remaining talent but face ongoing recruitment challenges competing with international remote work opportunities.

For AI implementation, this workforce situation suggests several strategic approaches: leveraging cloud-based AI services that minimize local technical maintenance requirements, establishing partnerships with international technology providers who can supplement local technical capacity, and investing in rapid training programs to upskill remaining health IT staff in AI system administration.

Training and Change Management Requirements #

Successful AI implementation requires more than technical deployment—it demands systematic change management and clinical staff training. Ukrainian medical education has historically included limited exposure to AI and machine learning concepts, meaning most practicing radiologists have minimal familiarity with AI diagnostic tools.

Recommended training approaches include:

• Hands-on workshops: Practical sessions where radiologists interact with AI systems on familiar case types
• Case-based learning: Review of AI-assisted diagnoses compared with expert consensus
• Error analysis training: Understanding AI system limitations and failure modes
• Workflow integration guidance: Best practices for incorporating AI recommendations into existing diagnostic processes
• Continuing education credits: Formal recognition of AI training to incentivize participation

10. Implementation Recommendations #

Based on the analysis of Ukraine’s healthcare infrastructure and AI readiness, the following phased implementation approach is recommended:

Phase	Timeline	Action	Rationale
Immediate	0-3 months	Map surviving MIS provider APIs	Establish integration foundation with viable systems
Short-term	3-9 months	Pilot with Lviv/Odesa centers	Leverage existing telemedicine infrastructure in stable regions
Medium-term	9-18 months	NHSU partnership proposal	Single-payer system enables system-wide deployment decisions
Long-term	18-36 months	Training data acquisition from EHS	Ukrainian-specific model development with local pathology patterns

11. Conclusions #

Ukraine’s healthcare system presents a complex but ultimately favorable landscape for AI-powered medical imaging deployment. The two-level EHS architecture provides standardized data infrastructure, the consolidated MIS ecosystem simplifies integration requirements, and the urgent need for diagnostic capacity amplifies adoption incentives.

Key findings from this analysis include:

• Medical imaging equipment density (~1.0-1.5 CT, ~0.5-0.8 MRI per 100K) falls significantly below EU standards, creating strong demand for AI-assisted efficiency improvements
• The eHealth system’s 36 million registered patients and 1.6 billion records provide substantial training data for Ukrainian-specific models
• Telemedicine surge during conflict has established infrastructure supporting remote AI-assisted diagnostics
• MIS consolidation from 40+ to 4-6 providers, while resulting from crisis, simplifies technical integration
• Private sector leadership in telemedicine suggests initial AI deployment should target private facilities

The path forward requires careful navigation of ongoing conflict impacts while leveraging Ukraine’s unexpected advantages in digital health infrastructure maturity and adoption urgency.

---