Abstract:Brain extraction is a critical pre-processing step in brain magnetic resonance imaging (MRI) analytical pipelines. In rodents, this is often achieved by manually editing brain masks slice-by-slice, a time-consuming task where workloads increase with higher spatial resolution datasets. We recently demonstrated successful automatic brain extraction via a deep-learning-based framework, U-Net, using 2D convolutions. However, such an approach cannot make use of the rich 3D spatial-context information from volumetric MRI data. In this study, we advanced our previously proposed U-Net architecture by replacing all 2D operations with their 3D counterparts and created a 3D U-Net framework. We trained and validated our model using a recently released CAMRI rat brain database acquired at isotropic spatial resolution, including T2-weighted turbo-spin-echo structural MRI and T2*-weighted echo-planar-imaging functional MRI. The performance of our 3D U-Net model was compared with existing rodent brain extraction tools, including Rapid Automatic Tissue Segmentation, Pulse-Coupled Neural Network, SHape descriptor selected External Regions after Morphologically filtering, and our previously proposed 2D U-Net model. 3D U-Net demonstrated superior performance in Dice, Jaccard, center-of-mass distance, Hausdorff distance, and sensitivity. Additionally, we demonstrated the reliability of 3D U-Net under various noise levels, evaluated the optimal training sample sizes, and disseminated all source codes publicly, with a hope that this approach will benefit rodent MRI research community. Significant Methodological Contribution: We proposed a deep-learning-based framework to automatically identify the rodent brain boundaries in MRI. With a fully 3D convolutional network model, 3D U-Net, our proposed method demonstrated improved performance compared to current automatic brain extraction methods, as shown in several qualitative metrics (Dice, Jaccard, PPV, SEN, and Hausdorff). We trust that this tool will avoid human bias and streamline pre-processing steps during 3D high resolution rodent brain MRI data analysis. The software developed herein has been disseminated freely to the community.

Fast and Lightweight Network Improves Serial Brain Section Stitching

Untrained Neural Network Enabling Fast and Universal Structured-Illumination Microscopy

Scalable Volumetric Imaging for Ultrahigh-Speed Brain Mapping at Synaptic Resolution

Semi-Automatic Neuron Tracing in Electron Microscope Images: Preliminary Results

Deep learning-Based 3D inpainting of brain MR images

Improved V-Net Based Image Segmentation for 3D Neuron Reconstruction.

Gossamer: Scaling Image Processing and Reconstruction to Whole Brains

3D U-Net Improves Automatic Brain Extraction for Isotropic Rat Brain Magnetic Resonance Imaging Data

Brain MR Atlas Construction Using Symmetric Deep Neural Inpainting

Brain Image Segmentation for Ultrascale Neuron Reconstruction via an Adaptive Dual-Task Learning Network

Dense 4D nanoscale reconstruction of living brain tissue

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

Nested U-Net Architecture Based Image Segmentation for 3D Neuron Reconstruction

3D Ultrasonic Brain Imaging with Deep Learning Based on Fully Convolutional Networks

Neuronal Population Reconstruction From Ultra-Scale Optical Microscopy Images via Progressive Learning

A deep learning-based strategy for producing dense 3D segmentations from sparsely annotated 2D images

Super-resolution Segmentation Network for Reconstruction of Packed Neurites

3D Dilated Multi-Fiber Network for Real-time Brain Tumor Segmentation in MRI

Progressive Learning for Neuronal Population Reconstruction from Optical Microscopy Images

DeepVolume: Brain Structure and Spatial Connection-Aware Network for Brain MRI Super-Resolution

An efficient memory reserving-and-fading strategy for vector quantization based 3D brain segmentation and tumor extraction using an unsupervised deep learning network