Abstract:The understanding of long-range pixel–pixel dependencies plays a vital role in image segmentation. The use of a CNN plus an attention mechanism still has room for improvement, since existing transformer-based architectures require many thousands of annotated training samples to model long-range spatial dependencies. This paper presents a smooth attention branch (SAB), a novel architecture that simplifies the understanding of long-range pixel–pixel dependencies for biomedical image segmentation in small datasets. The SAB is essentially a modified attention operation that implements a subnetwork via reshaped feature maps instead of directly calculating a softmax value over the attention score for each input. The SAB fuses multilayer attentive feature maps to learn visual attention in multilevel features. We also introduce position blurring and inner cropping specifically for small-scale datasets to prevent overfitting. Furthermore, we redesign the skip pathway for the reduction of the semantic gap between every captured feature of the contracting and expansive path. We evaluate the architecture of U-Net with the SAB (SAB-Net) by comparing it with the original U-Net and widely used transformer-based models across multiple biomedical image segmentation tasks related to the Brain MRI, Heart MRI, Liver CT, Spleen CT, and Colonoscopy datasets. Our training set was made of random 100 images of the original training set, since our goal was to adopt attention mechanisms for biomedical image segmentation tasks with small-scale labeled data. An ablation study conducted on the brain MRI test set demonstrated that every proposed method achieved an improvement in biomedical image segmentation. Integrating the proposed methods helped the resulting models consistently achieve outstanding performance on the above five biomedical segmentation tasks. In particular, the proposed method with U-Net improved its segmentation performance over that of the original U-Net by 13.76% on the Brain MRI dataset. We proposed several novel methods to address the need for modeling long-range pixel–pixel dependencies in small-scale biomedical image segmentation. The experimental results illustrated that each method could improve the medical image segmentation accuracy to various degrees. Moreover, SAB-Net, which integrated all proposed methods, consistently achieved outstanding performance on the five biomedical segmentation tasks.

Attention-modulated Multi-Branch Convolutional Neural Networks for Neonatal Brain Tissue Segmentation

Automated 3D Fetal Brain Segmentation Using an Optimized Deep Learning Approach

Attention Based Multi-Scale Nested Network for Biomedical Image Segmentation

Multi-task Learning for Neonatal Brain Segmentation Using 3D Dense-Unet with Dense Attention Guided by Geodesic Distance

MG-Net: A fetal brain tissue segmentation method based on multiscale feature fusion and graph convolution attention mechanisms

Deep Learning-Based Multiclass Brain Tissue Segmentation in Fetal MRIs

3D‐MASNet: 3D mixed‐scale asymmetric convolutional segmentation network for 6‐month‐old infant brain MR images

Isointense Infant Brain Segmentation with a Hyper-Dense Connected Convolutional Neural Network.

Multi Stream 3D Hyper-Densely Connected Network for Multi Modality Isointense Infant Brain MRI Segmentation

From Neonatal To Adult Brain Mr Image Segmentation In A Few Seconds Using 3d-Like Fully Convolutional Network And Transfer Learning

Multi-scale Multi-hierarchy Attention Convolutional Neural Network for Fetal Brain Extraction

nnSegNeXt: A 3D Convolutional Network for Brain Tissue Segmentation Based on Quality Evaluation

TW-Net: Transformer Weighted Network for Neonatal Brain MRI Segmentation

APRNet: A 3D Anisotropic Pyramidal Reversible Network with Multi-Modal Cross-Dimension Attention for Brain Tissue Segmentation in MR Images.

Harmonized neonatal brain MR image segmentation model for cross-site datasets

AMCNet: Attention-Based Multiscale Convolutional Network for DCM MRI Segmentation

Automated neonatal nnU-Net brain MRI extractor trained on a large multi-institutional dataset

CNS - CycleGAN-Assisted Neonatal Segmentation Model for Cross-Datasets.

A Deep Spatial Context Guided Framework for Infant Brain Subcortical Segmentation.

Deep CNN ensembles and suggestive annotations for infant brain MRI segmentation

Attention Mechanism Trained with Small Datasets for Biomedical Image Segmentation