Master’s Thesis Defense of Hasan Faleh Hasan

The Master’s thesis defense of Hasan Faleh Hasan was held in the Department of Biomedical Engineering on Thursday, August 28, 2025, for his thesis entitled:

“Artificial Intelligence Model Design for Detection and Classification of Intracranial Hemorrhage”

The examination committee consisted of:

• Chairman: Prof. Ahmed Abdulsameea Abdulwahab – College of Engineering / Al-Nahrain University

• Member: Assoc. Prof. Basheera Mohammed Reda – College of Engineering / University of Baghdad

• Member: Asst. Prof. Iman Ghidban Khalil – College of Engineering / Al-Nahrain University

The thesis was supervised by:

• **Assoc. Prof. Hadeel Qasim Wadi – College of Engineering / Al-Nahrain University

The thesis was scientifically evaluated by:

• First Scientific Reviewer: Prof. Muhannad Kazem Saber – College of Engineering Al-Khwarizmi / University of Baghdad

• Second Scientific Reviewer: Assoc. Prof. Hiba Zuhair Zidan – College of Information Engineering / Al-Nahrain University

The thesis was linguistically reviewed by:

• **Asst. Prof. Safa Laith Keilan – College of Engineering / Al-Nahrain University

Study Objective:
This research aims to design an intelligent system based on deep learning for the accurate detection and classification of intracranial hemorrhage types in CT scans, serving as a supportive tool for physicians to speed up and improve diagnosis.

Methodology:
The study involved the development and evaluation of five advanced AI models (DenseNet201, InceptionResNetV2, U-Net, U-Net-MobileNet, and U-Net-Xception) using Python. The models were trained on a large, standardized dataset (RSNA) containing thousands of images, and further tested on real hospital data to validate their effectiveness in real-world scenarios.

Results:
The proposed models achieved very high accuracy and reliability in detecting different types of hemorrhage (intraparenchymal, intraventricular, and subdural). DenseNet201 outperformed others in classification tasks, while U-Net excelled in localizing hemorrhages by generating precise heatmaps that highlight the affected regions for physicians.

Conclusion:
This study confirms that artificial intelligence tools can provide a practical and rapid solution that enhances physicians’ ability to timely diagnose critical brain injuries, thereby saving lives and improving healthcare outcomes.

The thesis was accepted as it met all the requirements for the Master’s degree.