Back to search

University of Ioannina

Βελτιστοποιημένες τεχνικές μηχανικής όρασης με χρήση συνελικτικών νευρωνικών δικτύων

Abstract

dc:description

This doctoral dissertation presents the development, implementation, and clinical evaluation of the Deep-in-Biopsies platform, a comprehensive system for automated histopathological image analysis using advanced deep learning techniques. The research focuses on addressing critical challenges in digital pathology, including diagnostic subjectivity, lack of integrated analysis tools, and the need for reliable, reproducible results. The primary contribution of this thesis is the creation of an innovative platform that unifies all phases of digital histopathology - from data annotation to training and evaluation of artificial intelligence models. The platform features a modular microservices architecture, dynamic management of multiple deep learning models, and flexible adaptation to different clinical scenarios. The research work includes the development and publication of five scientific papers in international conferences and journals, presenting experimental results across various tissue types (liver, kidney, intestine, breast). Specifically, methods were developed and evaluated for quantifying pancreatic steatosis, analyzing hepatic fibrosis, classifying renal tumors according to Fuhrman grading, and multi-organ cancer cell detection through the PanNuke dataset. Clinical validation across seven medical institutions demonstrated that the platform achieves diagnostic accuracy of 89-94% in various applications, reducing analysis time by 65% compared to traditional methods. The integration of explainable AI models and user-friendly interface facilitated acceptance by the medical community, with 87% of users reporting improved diagnostic confidence. The dissertation significantly contributes to advancing digital pathology by offering a practical, scalable, and clinically applicable solution that bridges the gap between research innovation and clinical practice.

Degree

thesis:*
Grantor dc:publisher
University of Ioannina
Year dc:date
2025

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Tsakai, Ontise
  • Τσακάι, Οντισέ

Subjects

dc:subject × 16

Rights

Language dc:language
gre

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/60715

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Tsakai, Ontise; Τσακάι, Οντισέ. Βελτιστοποιημένες τεχνικές μηχανικής όρασης με χρήση συνελικτικών νευρωνικών δικτύων. University of Ioannina, 2025. http://hdl.handle.net/10442/hedi/60715