3D fully convolutional networks for subcortical segmentation in MRI: A large-scale study

Abstract

This study investigates a 3D and fully convolutional neural network (CNN) for subcortical brain structure segmentation in MRI. 3D CNN architectures have been generally avoided due to their computational and memory requirements during inference. We address the problem via small kernels, allowing deeper architectures. We further model both local and global context by embedding intermediate-layer outputs in the final prediction, which encourages consistency between features extracted at different scales and embeds fine-grained information directly in the segmentation process. Our model is efficiently trained end-to-end on a graphics processing unit (GPU), in a single stage, exploiting the dense inference capabilities of fully CNNs. We performed comprehensive experiments over two publicly available datasets. First, we demonstrate a state-of-the-art performance on the ISBR dataset. Then, we report a {\em large-scale} multi-site evaluation over 1112 unregistered subject datasets acquired from 17 different sites (ABIDE dataset), with ages ranging from 7 to 64 years, showing that our method is robust to various acquisition protocols, demographics and clinical factors. Our method yielded segmentations that are highly consistent with a standard atlas-based approach, while running in a fraction of the time needed by atlas-based methods and avoiding registration/normalization steps. This makes it convenient for massive multi-site neuroanatomical imaging studies. To the best of our knowledge, our work is the first to study subcortical structure segmentation on such large-scale and heterogeneous data.

Accepted in the special issue of Neuroimage: "Brain Segmentation and Parcellation"

References in corpus (4)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Fully Convolutional Networks for Semantic Segmentation
• Automatic segmentation of MR brain images with a convolutional neural network
• Deep Neural Networks for Anatomical Brain Segmentation

Cited by in corpus (57)

• A review of deep learning in medical imaging: Imaging traits, technology trends, case studies with progress highlights, and future promises
• FastSurfer -- A fast and accurate deep learning based neuroimaging pipeline
• Boundary loss for highly unbalanced segmentation
• RA-UNet: A hybrid deep attention-aware network to extract liver and tumor in CT scans
• Deep convolutional neural networks for brain image analysis on magnetic resonance imaging: a review
• Constrained-CNN losses for weakly supervised segmentation
• Applications of Deep Learning Techniques for Automated Multiple Sclerosis Detection Using Magnetic Resonance Imaging: A Review
• Automatic autism spectrum disorder detection using artificial intelligence methods with MRI neuroimaging: A review
• Few-shot 3D Multi-modal Medical Image Segmentation using Generative Adversarial Learning
• Automated sub-cortical brain structure segmentation combining spatial and deep convolutional features
• Deep and Statistical Learning in Biomedical Imaging: State of the Art in 3D MRI Brain Tumor Segmentation
• Multi-region segmentation of bladder cancer structures in MRI with progressive dilated convolutional networks
• Bayesian convolutional neural network based MRI brain extraction on nonhuman primates
• Semi-supervised few-shot learning for medical image segmentation
• Brain segmentation based on multi-atlas guided 3D fully convolutional network ensembles
• Do We Really Need Dice? The Hidden Region-Size Biases of Segmentation Losses
• Quantitative analysis of patch-based fully convolutional neural networks for tissue segmentation on brain magnetic resonance imaging
• Deep Weakly-Supervised Learning Methods for Classification and Localization in Histology Images: A Survey
• Can autism be diagnosed with AI?
• Deep Learning for Neuroimaging-based Diagnosis and Rehabilitation of Autism Spectrum Disorder: A Review
• End-to-end semantic segmentation of personalized deep brain structures for non-invasive brain stimulation
• Topological Data Analysis for Multivariate Time Series Data
• Constrained domain adaptation for Image segmentation
• Segmentation with mixed supervision: Confidence maximization helps knowledge distillation
• Development of accurate human head models for personalized electromagnetic dosimetry using deep learning
• DARTS: DenseUnet-based Automatic Rapid Tool for brain Segmentation
• Whole Brain Segmentation with Full Volume Neural Network
• Bounding boxes for weakly supervised segmentation: Global constraints get close to full supervision
• Anisotropic mesh adaptation for region-based segmentation accounting for image spatial information
• Multi-Task Learning for Arousal and Sleep Stage Detection Using Fully Convolutional Networks
• Deep Co-Training for Semi-Supervised Image Segmentation
• Almost instant brain atlas segmentation for large-scale studies
• FD-FCN: 3D Fully Dense and Fully Convolutional Network for Semantic Segmentation of Brain Anatomy
• Nonlinear Markov Random Fields Learned via Backpropagation
• A Bayesian Hierarchical Network for Combining Heterogeneous Data Sources in Medical Diagnoses
• A deep learning classification scheme based on augmented-enhanced features to segment organs at risk on the optic region in brain cancer patients
• Harmonizing Flows: Leveraging normalizing flows for unsupervised and source-free MRI harmonization
• Accurate Automatic Segmentation of Amygdala Subnuclei and Modeling of Uncertainty via Bayesian Fully Convolutional Neural Network
• Graph Convolutions on Spectral Embeddings: Learning of Cortical Surface Data
• An Auto-Encoder Strategy for Adaptive Image Segmentation
• Privacy Preserving for Medical Image Analysis via Non-Linear Deformation Proxy
• Isointense Infant Brain Segmentation with a Hyper-dense Connected Convolutional Neural Network
• Label merge-and-split: A graph-colouring approach for memory-efficient brain parcellation
• Discretely-constrained deep network for weakly supervised segmentation
• Knowing what you know in brain segmentation using Bayesian deep neural networks
• On Direct Distribution Matching for Adapting Segmentation Networks
• Segmentation-by-Detection: A Cascade Network for Volumetric Medical Image Segmentation
• Spectral Graph Transformer Networks for Brain Surface Parcellation
• Learnable Pooling in Graph Convolution Networks for Brain Surface Analysis
• Bladder segmentation based on deep learning approaches: current limitations and lessons
• Unbiased Shape Compactness for Segmentation
• An ILP Solver for Multi-label MRFs with Connectivity Constraints
• DOPE: Distributed Optimization for Pairwise Energies
• Beyond one-hot encoding? Journey into compact encoding for large multi-class segmentation
• Computational Anatomy for Multi-Organ Analysis in Medical Imaging: A Review
• Deep Cerebellar Nuclei Segmentation via Semi-Supervised Deep Context-Aware Learning from 7T Diffusion MRI
• SUSAN: Segment Unannotated image Structure using Adversarial Network