Brain image segmentation is crucial for diagnosing strokes accurately. This study focused on using CT scan images to separate stroke objects from other structures. They applied a method called thresholding, which involves setting boundaries to distinguish different parts of the image. By using global binary and Otsu thresholding techniques, they were able to improve image quality significantly. The best results were achieved with Otsu thresholding, specifically by setting a lower threshold value of ≤170. This method helped identify stroke objects with high precision, making it easier to analyze and diagnose different types of strokes.