A Comprehensive Review of Deep Learning-Based Image Segmentation Architectures Applied to Various Diseases
A Comprehensive Review of Deep Learning-Based Image Segmentation Architectures Applied to Various Diseases

Author(s): Hafsa Laci, Kozeta Sevrani
Subject(s): Social Sciences
Published by: Udruženje ekonomista i menadžera Balkana
Summary/Abstract: Image segmentation plays a significant role in facilitating the arduous process of medical image analysis. There are numerous ways to perform image segmentation, but deep learning architectures have brought about a revolution in this field by automating it and improving the accuracy of the results. However, due to the complex nature of X-ray, MRI, or Ultrasound medical image modalities used for diagnosis, selecting the appropriate segmentation architecture becomes a challenging task. This review follows a systematic methodology to screen the literature and explore the available deep learning-based segmentation architectures applied across different diseases. It aims to contribute to existing research by identifying if in the state-of-art exists any approach that can be adaptive for a broader range of diseases. Furthermore, it seeks to evaluate computational and performance efficiency when there is evidence.

• Details
• Contents