Prof. A. Amalin Prince

Instituto Politécnico Nacional, Mexico

Speech Title: Multi-Objective Contrast–Noise Optimization in Medical Images with Preference Articulation

Abstract: In medical image enhancement, preserving image quality is essential, as excessive contrast amplification may introduce noise and artifacts that compromise diagnostic reliability. This paper proposes a preference articulation multi-objective optimization framework designed to accurately approximate predefined image quality levels while achieving superior Pareto front quality. The framework integrates: (i) a reference-based selection mechanism which incorporates a set of target references representing specific image quality levels, (ii) an adaptive reference direction strategy that dynamically redistributes search directions, and (iii) a local Newton--Raphson refinement mechanism that accelerates convergence toward the target references. Experimental results on a diverse set of medical images show that the proposed method outperforms MOEA/D and baseline variants, as reflected by higher success rates in target approximation as well as better hypervolume and spacing values. Statistical analysis confirms consistent and significant improvements across the experiments. Visual results further illustrate the effectiveness of the approach, showing enhanced contrast and improved structural visibility, while highlighting the trade-off between contrast enhancement and noise amplification.

Biography: DANIEL MOLINA-PÉREZ received the Ph.D. degree in Robotic and Mechatronic Systems from the Centro de Innovación y Desarrollo Tecnológico en Cómputo (CIDETEC), Instituto Politécnico Nacional, Mexico City, in 2024. He is currently a Professor-Researcher with the Escuela Superior de Cómputo (ESCOM), Instituto Politécnico Nacional. He is a member of the Artificial Intelligence and Data Science Network of the Instituto Politécnico Nacional. His research interests include evolutionary computation, multi-objective optimization, mixed-integer nonlinear programming, image enhancement, and hybrid optimization methods for solving complex engineering problems. Dr. Molina-Pérez has authored and co-authored multiple articles in indexed journals, including Swarm and Evolutionary Computation, PeerJ Computer Science, and Mathematical and Computational Applications, and has presented his work at leading international conferences such as the IEEE Congress on Evolutionary Computation (CEC). He received the Best Paper Award at the Mexican International Conference on Artificial Intelligence in 2022 and was recognized with the Outstanding Academic Performance Award at the doctoral level, reflecting his academic excellence and research contributions.

Institute of Mathematics of the Ministry of Science and Education, Republic of Azerbaijan

Speech Title: Cognitive Fuzzy Models for Confidentiality Assessment in Electronic Document Management Systems

Abstract: With the increasing digitalization of organizational workflows, reliable confidentiality assessment in electronic document management systems (EDMS) remains a challenging task. This paper proposes a cognitive fuzzy approach for intelligent classification and evaluation of document confidentiality under uncertainty and incomplete information. The proposed method integrates fuzzy logic, expert knowledge processing, and neuro-fuzzy modeling to support adaptive decision-making. Fuzzy inference is used to formalize subjective assessments, while neural models enable learning of complex cause-effect relations. The developed framework improves the accuracy and consistency of confidentiality evaluation and reduces the influence of the human factor. Experimental results obtained from a real electronic document management system demonstrate the effectiveness of the proposed approach in enhancing classification performance.

Biography: Prof. Ramin Rzayev is the Head of the Laboratory "Application of Artificial Intelligence Methods in Control Systems" at the Institute of Mathematics of the Ministry of Science and Education of the Republic of Azerbaijan. His research activities focus on the application of artificial intelligence methods to complex control and decision-making problems.
His main research interests include AI-based decision support under uncertainty, modeling and identification of complex weakly structured systems, neuro-fuzzy methods, intelligent information systems, digital ecosystems, and digital twin technologies. He also works on the development of AI-based methods for monitoring, prediction, optimization, and decision-making in complex systems, with applications in transportation, environmental monitoring, urban infrastructure, and other socio-technical systems.
He has authored and co-authored numerous scientific publications and has supervised research projects and doctoral studies in the fields of system analysis, artificial intelligence, information processing, intelligent systems, and applied mathematical modeling.

 

University of Canberra, Australia

Speech Title: Exploratory Frameworks for Unified Multimodal Learning for Medical Image Translation

Abstract: Medical image translation is an important research area within multimodal artificial intelligence, enabling cross-modality synthesis to reduce scan acquisition costs and enhance clinical analysis. While traditional image-to-image translation architectures such as Pix2Pix and CycleGAN focus strictly on unimodal image data, emerging multimodal generative concepts suggest a transition toward more integrated, context-aware frameworks. This paper presents an exploratory experimental investigation of three prototype lightweight architectures for T1-to-T2 brain MRI translation using the publicly available IXI dataset: (1) Model A, which implements a structured multimodal conditional translation model using synthetically simulated patient demographics and acquisition hospital site parameters; (2) Model B, which integrates text-conditioned semantic embeddings generated from BioBERT text prompts describing imaging characteristics; and (3) Model C, which utilizes a cyclic multimodal feedback translation loop in a bidirectional unpaired translation style. The models are evaluated quantitatively using Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index Measure (SSIM), and Mean Absolute Error (MAE) against four standard baselines: standard image-only Pix2Pix, standard CycleGAN, U-Net regression (L1 loss only), and Attention U-Net. Experimental results demonstrate that while the prototype architectures learn coarse anatomical relationships, the synthesized outputs exhibit visible smoothing and transpose-convolution checkerboard artifacts. The work is positioned as an exploratory proof-of-concept investigation into lightweight open-source multimodal translation mechanisms rather than a clinically deployable synthesis system, establishing a reproducible and statistically rigorous experimental foundation for future medical generative AI research.

Biography: Dr. Girija Chetty has a Bachelors and Masters degree in Electrical Engineering and Computer Science from India, and PhD in Information Sciences and Engineering from Australia. She has more than 40 years of experience in Industry, Research and Teaching from Universities and Research and Development Organisations from India and Australia, and has held several leadership positions including Head of Software Engineering and Computer Science, Program Director of ITS courses, and Course Director for Master of Computing and Information Technology Courses. Currently, she is a Full Professor in Computing and Information Technology at School of Information Technology and Systems, at University of Canberra, Australia, and leads a research group with several PhD students, Post Docs, research assistants and regular International and National visiting researchers. She is a Senior Member of IEEE, USA, and senior member of Australian Computer Society, and ACM member, and her research interests are in the area of multimodal systems, computer vision, pattern recognition, data mining, and medical image computing. She has published extensively with more than 200 fully refereed publications in several invited book chapters, edited books, high quality conference and journals, and she is in the editorial boards, technical review committees and regular reviewer for several IEEE, Elsevier and IET journals in the area related to her research interests. She is highly interested in seeking wide and interdisciplinary collaborations, research scholars and visitors in her research group.