Abstract
Manual methods for the task of brain tumor identification and analysis require a high amount of domain knowledge, are expensive, time consuming, and at times, inaccurate. Artificial intelligence plays a major role here as it can accomplish the same task via MRI segmentation in a less time-intensive manner while providing high accuracy. In this chapter, we study various semantic segmentation network architectures such as U-net, ResNet, and PSPNet used for the purpose of tumor segmentation from 3-D MRIs. We compare and analyze various parameters (such as accuracy, loss, etc.) of each of these network architectures.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have