Modality specific U-Net variants for biomedical image segmentation: a survey

Abstract

With the advent of advancements in deep learning approaches, such as deep convolution neural network, residual neural network, adversarial network; U-Net architectures are most widely utilized in biomedical image segmentation to address the automation in identification and detection of the target regions or sub-regions. In recent studies, U-Net based approaches have illustrated state-of-the-art performance in different applications for the development of computer-aided diagnosis systems for early diagnosis and treatment of diseases such as brain tumor, lung cancer, alzheimer, breast cancer, etc., using various modalities. This article contributes in presenting the success of these approaches by describing the U-Net framework, followed by the comprehensive analysis of the U-Net variants by performing (1) inter-modality, and (2) intra-modality categorization to establish better insights into the associated challenges and solutions. Besides, this article also highlights the contribution of U-Net based frameworks in the ongoing pandemic, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) also known as COVID-19. Finally, the strengths and similarities of these U-Net variants are analysed along with the challenges involved in biomedical image segmentation to uncover promising future research directions in this area.

3D Capsule Networks for Brain Image Segmentation

“…[20][21][22][23] The U-Net has shown strong auto-segmentation accuracy across a variety of different imaging modalities and anatomic structures. [21][22][23][24][25][26] Figure 1.B shows the architecture of our 3D U-Net. The input image undergoes 64 convolutions (3×3×3) to generate 64 feature maps.…”

Covid-MANet: Multi-task attention network for explainable diagnosis and severity assessment of COVID-19 from CXR images

“…2 (d) displays the building block [43] of DenseUNet model. The training of segmentation models is achieved by hybrid segmentation loss function [45] and the best model selection by IoU and Dice coefficient metrics. The references for compilation of a larger dataset compared to other models specified in section 3.1 , containing images and labeled lung annotated masks for supervision.…”

“…For infection region segmentation, we adopted UNet based encoder-decoder model [42] with a dense backbone network. UNet model with DenseNet121 backbone acts as an infection segmentation model, which is compared with seven other models involving UNet, Attention-UNet, UNet++, R2UNet [45] by considering backbone architecture as VGG19 with UNet++, ResNet50, and DenseNet201 with UNet. Similar to classification, the infection segmentation model is tested on two scenarios as of classification.…”

SegChaNet: A Novel Model for Lung Cancer Segmentation in CT Scans

“…Researchers have proposed many medical image analysis methods in CT scans to segregate the lung parenchyma region automatically [ 6 ]. For example, authors in [ 7 ] describe signal thresholding strategies based on contrast information for most methods. Because of their lower densities than the rest of the body, the lung region looks darker when framed by a denser region.…”