Author, Institution: Domas Jonaitis, Kaunas University of Technology
Science area, field of science: Technological Sciences, Informatics Engineering, T007
Research supervisor: Prof. Dr. Vidas Raudonis (Kaunas University of Technology, Technological Sciences, Informatics Engineering, T007)
Dissertation Defence Board of Informatics Engineering Science Field:
Prof. Dr. Renaldas Urniežius (Kaunas University of Technology, Technological Sciences, Informatics Engineering, T007) – chairperson
Prof. Dr. Gintaras Dervinis (Kaunas University of Technology, Technological Sciences, Informatics Engineering, T007)
Prof. Dr. Reza Ghabcheloo (Tampere University, Finland, Technological Sciences, Informatics Engineering, T007)
Prof. Dr. Nikolaj Goranin (Vilnius Gediminas Technical University, Technological Sciences, Informatics Engineering, T007)
Prof. Dr. Aušra Saudargienė (Vytautas Magnus University, Natural Sciences, Informatics, N009)
Dissertation defence meeting will be at Rectorate Hall of Kaunas University of Technology (K. Donelaičio 73–402, Kaunas)
The doctoral dissertation and summary are available at the library of Kaunas University of Technology (Gedimino g. 50, Kaunas) and on the internet:
D. Jonaitis el. dissertation.pdf
© D. Jonaitis, 2026 “The text of the thesis may not be copied, distributed, published, made public, including by making it publicly available on computer networks (Internet), reproduced in any form or by any means, including, but not limited to, electronic, mechanical or other means. Pursuant to Article 25(1) of the Law on Copyright and Related Rights of the Republic of Lithuania, a person with a disability who has difficulties in reading a document of a thesis published on the Internet, and insofar as this is justified by a particular disability, shall request that the document be made available in an alternative form by e-mail to doktorantura@ktu.lt.”
Annotation: This dissertation addresses the problem of subjective human embryo quality assessment during in vitro fertilization (IVF). Current clinical practice relies on single-focal-plane images, resulting in the loss of spatial biological information. Furthermore, existing artificial intelligence (AI) models operate as uninterpretable systems, limiting their clinical adoption. The aim of this research is to develop an automated, objective, and interpretable computational intelligence framework for embryo grading. The proposed system consists of three components. First, a U-Net-based deep learning architecture executes multi-focus image fusion, compressing 7-plane Z-stacks into a single 2D representation that preserves 3D morphology. Second, these representations are classified using hierarchical Vision Transformers (Swin-Tiny). Third, to ensure decision transparency, a Retrieval-Augmented Generation (RAG) framework is integrated, linking the classifier’s output with official clinical guidelines (ESHRE/ALPHA consensus). Experimental results demonstrate that multi-focus fusion improved classification accuracy by nearly 10%, achieving a 94.12% overall accuracy (0.995 macro AUC). By applying local (L1) pruning strategies, the model’s size was reduced while maintaining 93.02% accuracy at 20% sparsity. The inherently low computational cost of 4.51 GFLOPs and the ~6.6 s generative subsystem latency prove the feasibility of deploying the system on resource-constrained clinical hardware.