Abstract:The accurate segmentation of brain tumor is significant in clinical practice. Convolutional Neural Network (CNN)-based methods have made great progress in brain tumor segmentation due to powerful local modeling ability. However, brain tumors are frequently pattern-agnostic, i.e. variable in shape, size and location, which can not be effectively matched by traditional CNN-based methods with local and regular receptive fields. To address the above issues, we propose a shape-scale co-awareness network (S2CA-Net) for brain tumor segmentation, which can efficiently learn shape-aware and scale-aware features simultaneously to enhance pattern-agnostic representations. Primarily, three key components are proposed to accomplish the co-awareness of shape and scale. The Local-Global Scale Mixer (LGSM) decouples the extraction of local and global context by adopting the CNN-Former parallel structure, which contributes to obtaining finer hierarchical features. The Multi-level Context Aggregator (MCA) enriches the scale diversity of input patches by modeling global features across multiple receptive fields. The Multi-Scale Attentive Deformable Convolution (MS-ADC) learns the target deformation based on the multiscale inputs, which motivates the network to enforce feature constraints both in terms of scale and shape for optimal feature matching. Overall, LGSM and MCA focus on enhancing the scale-awareness of the network to cope with the size and location variations, while MS-ADC focuses on capturing deformation information for optimal shape matching. Finally, their effective integration prompts the network to perceive variations in shape and scale simultaneously, which can robustly tackle the variations in patterns of brain tumors. The experimental results on BraTS 2019, BraTS 2020, MSD BTS Task and BraTS2023-MEN show that S2CA-Net has superior overall performance in accuracy and efficiency compared to other state-of-the-art methods. Code: https://github.com/jiangyu945/S2CA-Net.

Quality-Aware Model Ensemble for Brain Tumor Segmentation

Tri-PANet: Triplanar Ensemble Network with Parallelly Aggregating Global and Local Information for Brain Tumor Segmentation

Shape-Scale Co-Awareness Network for 3D Brain Tumor Segmentation.

Model Ensemble for Brain Tumor Segmentation in Magnetic Resonance Imaging

Learning Contextual and Attentive Information for Brain Tumor Segmentation

Deep learning-based ensemble model for brain tumor segmentation using multi-parametric MR scans

Ensemble Outperforms Single Models in Brain Tumor Segmentation

Magnetic Resonance Imaging Feature-Based Subtyping and Model Ensemble for Enhanced Brain Tumor Segmentation

Asymmetric Ensemble of Asymmetric U-Net Models for Brain Tumor Segmentation With Uncertainty Estimation

Automated ensemble method for pediatric brain tumor segmentation

Brain Tumor Segmentation Using an Ensemble of 3D U-Nets and Overall Survival Prediction Using Radiomic Features

Ensemble Learning with Residual Transformer for Brain Tumor Segmentation

Ensemble Technique for Brain Tumor Patient Survival Prediction

Automated Ensemble-Based Segmentation of Adult Brain Tumors: A Novel Approach Using the BraTS AFRICA Challenge Data

Brain tumor segmentation with self-ensembled, deeply-supervised 3D U-net neural networks: a BraTS 2020 challenge solution

XAI-MRI: An Ensemble Dual-Modality Approach for 3D Brain Tumor Segmentation Using Magnetic Resonance Imaging

A Deep-Learning Model with Learnable Group Convolution and Deep Supervision for Brain Tumor Segmentation

Multi-Scale Masked 3-D U-Net For Brain Tumor Segmentation

Automatic Brain Tumor Segmentation with Scale Attention Network

Attention-Enhanced Hybrid Feature Aggregation Network for 3D Brain Tumor Segmentation

Modality-Pairing Learning for Brain Tumor Segmentation