Abstract:Brain tumor segmentation from Magnetic Resonance Images (MRI) presents significant challenges due to the complex nature of brain tumor tissues. This complexity makes distinguishing tumor tissues from healthy tissues difficult, mainly when radiologists perform manual segmentation. Reliable and accurate segmentation is crucial for effective tumor grading and treatment planning. In this paper, we proposed a novel ensemble dual-modality approach for 3D brain tumor segmentation using MRI. Initially, individual U-Net models are trained and evaluated on single MRI modalities (T1, T2, T1ce, and FLAIR) to establish each modality's performance. Subsequently, we trained U-net models using combinations of the best-performing modalities to exploit the complementary information and improve segmentation accuracy. The most effective dual-modality combinations are then integrated into an ensemble model, designed to capture the strengths of each modality combination. Finally, we suggested the ensemble dual-modality by combining the best-performing two pre-trained dual-modalities models to enhance segmentation performance. Experimental results demonstrate the ensemble dual-modality model significantly improved segmentation performance, achieving a Dice Coefficient of 97.73% and a Mean IoU of 60.08% on the BraTS2020 dataset. Our results demonstrate that this ensemble approach outperforms traditional single-modality and dual-modality models, providing a more robust and accurate method for brain tumor segmentation. This study underscores the potential of ensemble dual-modality models to enhance the precision and reliability of MRI-based brain tumor segmentation. Our code publicly available at: https://github.com/Ahmeed-Suliman-Farhan/Ensemble-Dual-Modality-Approach .

AssemblyNet: A large ensemble of CNNs for 3D Whole Brain MRI Segmentation

AssemblyNet: A Novel Deep Decision-Making Process for Whole Brain MRI Segmentation

DeU-Net 2.0: Enhanced deformable U-Net for 3D cardiac cine MRI segmentation

Fast and Lightweight Network Improves Serial Brain Section Stitching

Deep CNN ensembles and suggestive annotations for infant brain MRI segmentation

Region-based U-net for accelerated training and enhanced precision in deep brain segmentation

Deep 3D Neural Network for Brain Structures Segmentation Using Self-Attention Modules in MRI Images

DAUnet: A U-shaped network combining deep supervision and attention for brain tumor segmentation

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

Brain segmentation based on multi-atlas guided 3D fully convolutional network ensembles

ACEnet: Anatomical Context-Encoding Network for Neuroanatomy Segmentation

Dimensionality Reduction Hybrid U-Net for Brain Extraction in Magnetic Resonance Imaging

DVNet: A Memory-Efficient Three-Dimensional CNN for Large-Scale Neurovascular Reconstruction

DeepNAT: Deep Convolutional Neural Network for Segmenting Neuroanatomy

Hough-CNN: Deep learning for segmentation of deep brain regions in MRI and ultrasound

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

Automated joint skull-stripping and segmentation with Multi-Task U-Net in large mouse brain MRI databases

Whole brain segmentation with full volume neural network

Multimodal CNN Networks for Brain Tumor Segmentation in MRI: A BraTS 2022 Challenge Solution

Model Ensemble for Brain Tumor Segmentation in Magnetic Resonance Imaging

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