Title Page
Contents
Ⅰ. Introduction 5
1. Machine Learning and Neural Network 5
2. Machine Learning in Neurosurgery 8
3. Meningioma and Gamma Knife Radiosurgery 8
4. Radiation-induced Toxicities and Peri-tumoral Edema following SRS 9
5. Study Objective 13
Ⅱ. Materials and Methods 14
1. Patient Selection and Data Collection 14
2. Outcomes 16
3. Statistical Analysis 17
4. Machine Learning (ML) 17
1) Radiologic Image Acquisition and Preparation 18
2) Tumor Identifier Model and Generating Concatenated Tumor Images 19
3) Predictive Models for Post-GKS PTE 21
Ⅲ. Results 22
1. Patient Population and Tumor Characteristics 22
2. Primary and Secondary Outcomes 24
3. Performance of Predictive Models and Comparison with a Traditional Statistical Model 25
Ⅳ. Discussion 28
1. Machine Learning (ML) in the Field of Radiosurgery 28
2. Prediction of Post-GKS Edema 29
Ⅴ. Limitation 31
Ⅵ. Conclusion 32
Ⅶ. References 33
ABSTRACT 38