Biomedical Image Analysis by Using Heuristic Algorithms

Thesis Type: Postgraduate

Institution Of The Thesis: Erciyes University, Fen Bilimleri Enstitüsü, Turkey

Approval Date: 2018

Thesis Language: Turkish

Student: Hakan Duran

Supervisor: Mehmet Bahadır Çetinkaya


Computer based imaging and analysis techniques are frequently used for the diagnosis and treatment of medical diseases. Especially retinal diseases can successfully be detected and treated by using computer based analysis methods. Arteriolar stenosis, arteriolar dilatation and vascular bleeding due to retinal diseases give important information about the disease. Therefore, to detect structural changes in retinal images with high accuracy is extremely important for diagnosis and treatment of the disease.

In this thesis, firstly, the detection and analysis of retinal blood vessels is realized with vessel segmentation. Then, approaches have been improved for the aims of identification, clarification and pixel based area calculation of the new hemorrhagic regions arising from diabetes mellitus. Retinal image analyses are realized by using    K-means, artificial bee colony (ABC) and particle swarm optimization (PSO) algorithms. From the simulation results it is observed that both the K-means algorithm which is one of the classical approaches and the clustering based heuristic ABC and PSO algorithms can successfully be used in retinal image analysis.