Dense Volume-to-Volume Vascular Boundary Detection

By Jameson Tyler Merkow, Alison L. Marsden, David J. Kriegman, Zhuowen Tu. 2016

89 citations recorded by the source corpus.

In this work, we tackle the important problem of dense 3D volume labeling in medical imaging. We start by introducing HED-3D, a 3D extension of the state-of-the-art 2D edge detector (HED). Next, we develop a novel 3D-Convolutional Neural Network (CNN) architecture, I2I-3D, that predicts boundary location in volumetric data. Our fine-to-fine, deeply supervised framework addresses three critical issues to 3D boundary detection: (1) efficient, holistic, end-to-end volumetric label training and prediction (2) precise voxel-level prediction to capture fine scale structures prevalent in medical data and (3) directed multi-scale, multi-level feature learning. We evaluate our approaches on a dataset consisting of 93 medical image volumes with a wide variety of anatomical regions and vascular structures. We show that our deep learning approaches out-perform the current state-of-the-art in 3D vascular boundary detection (structured forests 3D), by a large margin, as well as HED applied to slices. Prediction takes about one minute on a typical \(512\,\times \,512\,\times \,512\) volume, when using GPU.

Subfields: Artificial Intelligence, Biomedical Engineering, Computer Vision and Pattern Recognition, Medical Imaging

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Biomedical Engineering

In Computer Vision and Pattern Recognition

In Medical Imaging

Citations in the corpus

Holistically-Nested Edge Detection (2015, background): “…response fusion [38, 30, 28], (3) engagement of different levels of visual perception [16, 25, 37, 15] such as mid-level Gestalt law information [7], (4) structural information [6] and context [36], (5) holistic image prediction [23], (6) 3D geometry [13], and (7) occlusion boundaries [14].”
Spatial aggregation of holistically‐nested convolutional neural networks for automated pancreas localization and segmentation☆ (2017, methodology): “, 96×96×48 [34], 160×160×72 [32] or 80×80×80 [33], which may cause the segmentation discontinuity or inconsistency at overlapped window boundaries.”
On the Compactness, Efficiency, and Representation of 3D Convolutional Networks: Brain Parcellation as a Pretext Task (2017, background): “Most of the existing network architectures follow a fully convolutional downsample-upsample pathway [11,4,15,3,16,13].”
Volumetric ConvNets with Mixed Residual Connections for Automated Prostate Segmentation from 3D MR Images (2017, methodology): “By borrowing the design principles of U-Net (Ronneberger, Fischer, and Brox 2015), some researchers proposed 3D fully convolutional networks for efficient segmentation tasks (Çiçek et al. 2016; Milletari, Navab, and Ahmadi 2016; Merkow et al. 2016).”
MultiResUNet : Rethinking the U-Net Architecture for Multimodal Biomedical Image Segmentation (2019, methodology): “U-Net and U-Net like models have been successfully used in segmenting biomedical images of neuronal structures [24], liver [25], skin lesion [26], colon histology [27], kidney [28], vascular boundary [29], lung nodule [30], prostate [31], etc. and the list goes on.”
Deep Learning Techniques for Medical Image Segmentation: Achievements and Challenges (2019, background): “Training a network with such data often leads to the trained network being biased toward the background and got trapped in local minima [51, 53].”
VoxResNet: Deep voxelwise residual networks for brain segmentation from 3D MR images (2017, unknown): “…other type of methods employed real 3D CNNs to detect or segment objects from volumetric data and demonstrated compelling performance (Dou et al.; Dou et al., 2016a, 2016b; Cicek et al., 2016; Yu et al., 2017; Merkow et al., 2015; Milletari et al., 2016; Kamnitsas et al., 2017; Chen et al., 2016a).”
Computer Methods and Programs in Biomedicine (unknown): “In [33] , the FCN approach is extended to perform 3D vascular segmentation in cardiovascular and pulmonary vessels in MRI and CT volumes, respectively.”
3D deeply supervised network for automated segmentation of volumetric medical images (2017, methodology): “…( Dou et al., 2016a ), which we substantially revised in this extended version, several 3D volume-to-volume segmentation networks have been proposed, including 3D U-Net ( Çiçek et al., 2016 ), V-Net ( Milletari et al., 2016 ), I2I-3D ( Merkow et al., 2016 ) and VoxResNet ( Chen et al., 2017 ).”
Deep Learning for Cardiac Image Segmentation: A Review (2019, methodology): “End-to-end CNNs With respect to the lumen or vessel wall segmentation, most deep learning based approaches use an end-to-end CNN segmentation scheme to predict dense segmentation probability maps (38, 122, 126, 205).”