Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Monika Simjanoska has a total of 92 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Multi-Agent Readiness Scoring Methodology in Bioinformatics Domain

    by Blagojche Gjorgjioski, Djansel Bukovec, Ivana Vichentijevikj, Ivan Kitanovski, Kostadin Mishev, Monika Simjanoska Misheva

    The emergence of Large Language Models (LLMs) has significantly advanced computational biology, yet their integration into autonomous, multi-agent systems (MASs) and clinical workflows remains challenging due to systemic architectural fragmentation. To quantify the operational readiness and regulatory compliance of bioinformatics LLMs, we developed the Multi-Agent Readiness Score (MARS), a standardized evaluation framework assessing models across four structural dimensions: Governance & Accessibility, Biological Competence, Technical Maturity, and Agentic Orchestration. The framework incorporates compliance criteria from the EU AI Act, HL7 FHIR, HL7 CDA, and MyHealth@EU standards. To empirically validate this domain-agnostic methodology, we applied it to a highly mature subset of the field: a diverse cohort of 43 prominent genomic LLMs. Our assessment revealed a severe, industry-wide readiness gap: the majority of models fell into “Not Suitable” or “Research Prototype” tiers, lacking essential technical interfaces, structured communication schemas, and provenance tracking. Furthermore, the data demonstrated a ’competence-readiness gap’, where models scale in biological predictive competence without corresponding improvements in engineering utility. The primary barrier to scalable bioinformatics AI is no longer biological competence, but operational and architectural incompatibility. By quantifying integration friction, MARS provides a crucial, reproducible metric to audit model maturity, guide system architecture, and ensure future models are structurally prepared for the rigorous regulatory demands of precision medicine workflows.

  • Open Access

    Exploring the Potential of Topological Data Analysis for Explainable Large Language Models: A Scoping Review

    by Petar Sekuloski, Dimitar Kitanovski, Igor Goshev, Kostadin Mishev, Monika Simjanoska Misheva, Vesna Dimitrievska Ristovska

    Large language models (LLMs) have become central to modern artificial intelligence, yet their internal decision-making processes remain difficult to interpret. As interest grows in making these models more transparent and reliable, topological data analysis (TDA) has emerged as a promising mathematical approach for exploring their structure. This scoping review maps the current landscape of research where TDA tools—such as persistent homology and Mapper—are used to examine LLM components like attention patterns, latent representations, and training dynamics. By analyzing topological features across layers and tasks, these methods provide new ways to understand how language models generalize, respond to unfamiliar inputs, and shift under fine-tuning. The review also considers how TDA-based techniques contribute to broader goals in interpretability and robustness, especially in detecting hallucinations, out-of-distribution behavior, and representational collapse. Overall, the findings suggest that TDA offers a rigorous and versatile framework for studying LLMs, helping researchers uncover deeper patterns in how these models learn and reason.


  • Open Access

    AI Act Compliance Within the MyHealth@EU Framework: Tutorial

    by Monika Simjanoska Misheva, Dragan Shahpaski, Jovana Dobreva, Djansel Bukovec, Blagojche Gjorgjioski, Marjan Nikolov, Dalibor Frtunikj, Petre Lameski, Azir Aliu, Kostadin Mishev, et al.

    The integration of artificial intelligence (AI) into clinical workflows is advancing even before full compliance with the European Union Cross-Border eHealth Network (MyHealth@EU) framework is achieved. While AI-based clinical decision support systems are automatically classified as high risk under the European Union’s AI Act, cross-border health data exchange must also satisfy MyHealth@EU interoperability requirements. This creates a dual-compliance challenge: vertical safety and ethics controls mandated by the AI Act and horizontal semantic transport requirements enforced through Open National Contact Point (OpenNCP) gateways, many of which are still maturing toward production readiness. This paper provides a practical, phase-oriented tutorial that enables developers and providers to embed AI Act safeguards before approaching MyHealth@EU interoperability tests. The goal is to show how AI-specific metadata can be included in the Health Level Seven International Clinical Document Architecture and Fast Healthcare Interoperability Resources messages without disrupting standard structures, ensuring both compliance and trustworthiness in AI-assisted clinical decisions. We systematically analyzed Regulation (EU) 2024/1689 (AI Act) and the OpenNCP technical specifications, extracting a harmonized set of overlapping obligations. The AI Act provisions on transparency, provenance, and robustness are mapped directly onto MyHealth@EU workflows, identifying the points where outgoing messages must record AI involvement, log provenance, and trigger validation. To operationalize this mapping, we propose a minimal extension set, covering AI contribution status, rationale, risk classification, and Annex IV documentation links, together with a phase-based compliance checklist that aligns AI Act controls with MyHealth@EU conformance steps. A simulated International Patient Summary transmission demonstrates how Clinical Document Architecture/Fast Healthcare Interoperability Resources extensions can annotate AI involvement, how OpenNCP processes such enriched payloads, and how clinicians in another member state view the result with backward compatibility preserved. We expand on security considerations (eg, Open Worldwide Application Security Project generative AI risks such as prompt injection and adversarial inputs), continuous postmarket risk assessment, monitoring, and alignment with MyHealth@EU’s incident aggregation system. Limitations reflect the immaturity of current infrastructures and regulations, with real-world validation pending the rollout of key dependencies. AI-enabled clinical software succeeds only when AI Act safeguards and MyHealth@EU interoperability rules are engineered together from day 0 . This tutorial provides developers with a forward-looking blueprint that reduces duplication of effort, streamlines conformance testing, and embeds compliance early. While the concept is still in its early phases of practice, it represents a necessary and worthwhile direction for ensuring that future AI-enabled clinical systems can meet both European Union regulatory requirements from day 1. risks such as prompt injection and adversarial inputs), continuous postmarket risk assessment, monitoring, and alignment with MyHealth@EU’s incident aggregation system. Limitations reflect the immaturity of current infrastructures and regulations, with real-world validation pending the rollout of key dependencies.


  • Open Access

    RAGCare-QA: A benchmark dataset for evaluating retrieval-augmented generation pipelines in theoretical medical knowledge

    by Jovana Dobreva, Ivana Karasmanakis, Filip Ivanisevic, Tadej Horvat, Dimitar Kitanovski, Matjaz Gams, Kostadin Mishev, Monika Simjanoska Misheva


  • Open Access

    Evaluating a Nationally Localized AI Chatbot for Personalized Primary Care Guidance: Insights from the HomeDOCtor Deployment in Slovenia

    by Matjaž Gams, Tadej Horvat, Žiga Kolar, Primož Kocuvan, Kostadin Mishev, Monika Simjanoska Misheva

    Background/Objectives: The demand for accessible and reliable digital health services has increased significantly in recent years, particularly in regions facing physician shortages. HomeDOCtor, a conversational AI platform developed in Slovenia, addresses this need with a nationally adapted architecture that combines retrieval-augmented generation (RAG) and a Redis-based vector database of curated medical guidelines. The objective of this study was to assess the performance and impact of HomeDOCtor in providing AI-powered healthcare assistance. Methods: HomeDOCtor is designed for human-centered communication and clinical relevance, supporting multilingual and multimedia citizen inputs while being available 24/7. It was tested using a set of 100 international clinical vignettes and 150 internal medicine exam questions from the University of Ljubljana to validate its clinical performance. Results: During its six-month nationwide deployment, HomeDOCtor received overwhelmingly positive user feedback with minimal criticism, and exceeded initial expectations, especially in light of widespread media narratives warning about the risks of AI. HomeDOCtor autonomously delivered localized, evidence-based guidance, including self-care instructions and referral suggestions, with average response times under three seconds. On international benchmarks, the system achieved ≥95% Top-1 diagnostic accuracy, comparable to leading medical AI platforms, and significantly outperformed stand-alone ChatGPT-4o in the national context (90.7% vs. 80.7%, p = 0.0135). Conclusions: Practically, HomeDOCtor eases the burden on healthcare professionals by providing citizens with 24/7 autonomous, personalized triage and self-care guidance for less complex medical issues, ensuring that these cases are self-managed efficiently. The system also identifies more serious cases that might otherwise be neglected, directing them to professionals for appropriate care. Theoretically, HomeDOCtor demonstrates that domain-specific, nationally adapted large language models can outperform general-purpose models. Methodologically, it offers a framework for integrating GDPR-compliant AI solutions in healthcare. These findings emphasize the value of localization in conversational AI and telemedicine solutions across diverse national contexts.


  • Open Access

    AI Act Compliance within the MyHealth@EU Framework: A Tutorial (Preprint)

    by Monika Simjanoska Misheva, Dragan Shahpaski, Jovana Dobreva, Djansel Bukovec, Blagojche Gjorgjioski, Marjan Nikolov, Dalibor Frtunikj, Petre Lameski, Azir Aliu, Kostadin Mishev, et al.

    BACKGROUND The integration of AI within clinical workflows is anticipated even prior to achieving full compliance with the MyHealth@EU framework. Myriad use cases have been presented internationally and garnered media attention. AI-based Clinical Decision Support Systems (CDSS) are automatically classified as high-risk under the EU AI Act, while cross-border data sharing in Europe must satisfy the MyHealth@EU interoperability rules. Development teams will inevitably face a dual-compliance challenge: vertical safety and ethics controls mandated by the AI Act, and horizontal semantic-transport requirements enforced through OpenNCP software’s gateways; some of which are yet to be brought to production readiness. OBJECTIVE The aim of the paper is to provide a practical, phase-oriented tutorial that enables developers and providers to embed AI Act controls prior to approaching the MyHealth@EU interoperability tests. The goal is to enable the inclusion of AI-specific tags in the HL7 CDA and HL7 FHIR messages without compromising the standard structure, while ensuring trustworthiness in the AI-aided clinical decision. METHODS We systematically analysed Regulation (EU) 2024/1689 (henceforth “AI Act”) and the MyHealth@EU/OpenNCP technical specifications, cross-referencing official guidance and semantic assets to extract a harmonised set of overlapping compliance obligations. RESULTS The mandated duties within the AI Act’s transparency, provenance, and robustness provisions are precisely overlaid onto the existing MyHealth@EU workflow. These define where any outgoing clinical message must record AI involvement, log the provenance of this involvement, and trigger validation of the information. A lean extension set adds AI contribution status, rationale, risk class, and an Annex IV (AI Act) link while remaining schema compliant. A phase-based checklist then threads every high-risk AI control giving developers one path to simultaneous AI-Act and MyHealth@EU alignment. CONCLUSIONS AI-enabled clinical software succeeds only when AI Act safeguards and MyHealth@EU interoperability rules are engineered together from ‘day zero’. The proposed checklist and lightweight CDA/FHIR extensions let developers embed, test, and prove high-risk AI compliance while injecting the necessary AI metadata directly into standard messages. This integrated blueprint cuts duplication of effort and delivers clinically trustworthy, AI-annotated reports that flow safely and instantly across national borders. CLINICALTRIAL N/A

  • Open Access

    A Unified Framework for Alzheimer’s Disease Knowledge Graphs: Architectures, Principles, and Clinical Translation

    by Jovana Dobreva, Monika Simjanoska Misheva, Kostadin Mishev, Dimitar Trajanov, Igor Mishkovski

    This review paper synthesizes the application of knowledge graphs (KGs) in Alzheimer’s disease (AD) research, based on two basic questions, as follows: what types of input data are available to construct these knowledge graphs, and what purpose the knowledge graph is intended to fulfill. We synthesize results from existing works to illustrate how diverse knowledge graph structures behave in different data availability settings with distinct application targets in AD research. By comparative analysis, we define the best methodology practices by data type (literature, structured databases, neuroimaging, and clinical records) and application of interest (drug repurposing, disease classification, mechanism discovery, and clinical decision support). From this analysis, we recommend AD-KG 2.0, which is a new framework that coalesces best practices into a unifying architecture with well-defined decision pathways for implementation. Our key contributions are as follows: (1) a dynamic adaptation mechanism that adapts methodological elements automatically according to both data availability and application objectives, (2) a specialized semantic alignment layer that harmonizes terminologies across biological scales, and (3) a multi-constraint optimization approach for knowledge graph building. The framework accommodates a variety of applications, including drug repurposing, patient stratification for precision medicine, disease progression modeling, and clinical decision support. Our system, with a decision tree structured and pipeline layered architecture, offers research precise directions on how to use knowledge graphs in AD research by aligning methodological choice decisions with respective data availability and application goals. We provide precise component designs and adaptation processes that deliver optimal performance across varying research and clinical settings. We conclude by addressing implementation challenges and future directions for translating knowledge graph technologies from research tool to clinical use, with a specific focus on interpretability, workflow integration, and regulatory matters.


  • Exploring ASR Models in Low-Resource Languages: Use-Case the Macedonian Language

    by Konstantin Bogdanoski, Kostadin Mishev, Monika Simjanoska, Dimitar Trajanov