CARDIA-X: Global Semantic Transition and Rough-Set Rules for Auditable Post Hoc Electrocardiographic Explainability
AI 7(9):376, 2026.
Pavlo Radiuk
Associate Professor, Department of Computer Science
Khmelnytskyi National University, Ukraine
I work on explainable artificial intelligence. My methods map what a deep network has learned onto the features that clinicians and engineers already use, so that its decisions can be checked and audited. I apply them in medical imaging, UAV inspection of energy infrastructure and urban traffic control, and I am a researcher in the Horizon Europe project U_CAN.
radiukp [at] khmnu.edu.ua CV Department profile Google Scholar ORCID GitHub
Research
-
Medical AI
Explanations for ECG, cardiac MRI and chest radiographs, evaluated against clinical expert judgement.
-
Energy and UAV inspection
Detection and criticality assessment of defects in wind turbines and photovoltaic plants from UAV imagery.
-
Smart mobility
Reinforcement-learning signal control and traffic-pattern recognition for urban mobility.
Selected publications
Full list-
Figure 1 of this article · CC BY 4.0 -
Figure 1 of this article · CC BY 4.0 Equivariant Transition Matrices for Explainable Deep Learning: A Lie Group Linearization Approach
Machine Learning and Knowledge Extraction 8(4):92, 2026.
-
Figure 1 of this article · CC BY 4.0 Criticality Assessment of Wind Turbine Defects via Multispectral UAV Fusion and Fuzzy Logic
Energies 18(17):4523, 2025.
-
Figure 8 of this article · CC BY 4.0 Hierarchical Deep Learning Model for Identifying Similar Targets in UAV Imagery
Drones 9(11):743, 2025.
-
Figure 9 of this article · CC BY 4.0 Intelligent Integrated System for Fruit Detection Using Multi-UAV Imaging and Deep Learning
Sensors 24(6):1913, 2024.
-
Figure 2 of this article · CC BY-NC-ND 4.0 Impact of Training Set Batch Size on the Performance of Convolutional Neural Networks for Diverse Datasets
Information Technology and Management Science 20(1), 2017.
† mentored junior researcher
News
- ExplAI-2026 workshop to be held on 13 November 2026
- Cardiac MRI audit paper presented at ICST-ODESA 2026
- CARDIA-X published in AI
- Weather-robust UAV localisation article online in the Journal of Intelligent & Robotic Systems
Collaboration
I welcome partners for Horizon Europe, MSCA and EuroHPC proposals in explainable AI, medical AI, UAV inspection and smart mobility. The partner kit has a capability statement, short biographies and my institution’s participant code.