3. Accurate segmentation of medical images is critical for d…

3. Accurate segmentation of medical images is critical for disease diagnosis and treatment planning. Different segmentation methods are used in medical imaging, such as identifying brain tumors in MRI scans. 3.1 How can you determine whether the noise present in an MRI image (Fig. 4) is Gaussian noise or salt-and-pepper noise? How does this noise affect medical image analysis, particularly segmentation? 3.2a If you are operating the MRI scanner, how can you improve the signal to noise ratio? List at least one method. 3.2b If you already acquired this image, before segmenting the brain region, preprocessing techniques can improve image quality.  b.1) List two denoising techniques that can enhance the segmentation of the MRI image. b.2) Comparing acquisition-based SNR improvement vs. post-processing denoising. If you have the option to choose between improving SNR during acquisition vs. applying post-processing denoising, which would be the better choice and why? 3.3 Please provide at least three image segmentation methods, excluding thresholding, that have been applied to medical images in the course. 3.4 After segmentation, some regions of the brain appear disconnected or contain small holes inside the segmented area. What image processing method could you use to fill in the missing areas or connect small hole regions.

2. If you have acquired a DICOM image, it is necessary to id…

2. If you have acquired a DICOM image, it is necessary to identify the most relevant areas of the image based on the anatomical structure. 2.1 Explain briefly the differences between DICOM and standard gray-level images. Additionally, if a rectangle ROI on a given image (Fig. 1) is p1= (82, 140), p2=(182,140), p3=(80, 280), and p4= (182, 280), the ROI focuses on the key anatomical structure (heart). If the image resolution is 1.5mm x 1.5mm, what is the length and width of the heart? 2.2 Assuming you already have an ROI in this image, and a point processing function , how would you modify the image in Figure 2? What is the maximum value of the modified image? Please note that is the original image and is the modified image. 2.3 If we apply the point processing function to the image, what would be the maximum value of the resulting image and why? 2.4 What value of ‘a’ in the point processing function