DeepOrgan: Multi-level Deep Convolutional Networks for Automated Pancreas Segmentation

By Holger R. Roth, Le Lu, Amal A. Farag, Hoo-Chang Shin, Jiamin Liu, Evrim B. Turkbey, Ronald M. Summers. 2015

870 citations recorded by the source corpus.

Automatic organ segmentation is an important yet challenging problem for medical image analysis. The pancreas is an abdominal organ with very high anatomical variability. This inhibits previous segmentation methods from achieving high accuracies, especially compared to other organs such as the liver, heart or kidneys. In this paper, we present a probabilistic bottom-up approach for pancreas segmentation in abdominal computed tomography CT scans, using multi-level deep convolutional networks ConvNets. We propose and evaluate several variations of deep ConvNets in the context of hierarchical, coarse-to-fine classification on image patches and regions, i.e. superpixels. We first present a dense labeling of local image patches via P-ConvNet and nearest neighbor fusion. Then we describe a regional ConvNet R1-ConvNet that samples a set of bounding boxes around each image superpixel at different scales of contexts in a "zoom-out" fashion. Our ConvNets learn to assign class probabilities for each superpixel region of being pancreas. Last, we study a stacked R2-ConvNet leveraging the joint space of CT intensities and the P-ConvNet dense probability maps. Both 3D Gaussian smoothing and 2D conditional random fields are exploited as structured predictions for post-processing. We evaluate on CT images of [InlineEquation not available: see fulltext.] patients in 4-fold cross-validation. We achieve a Dice Similarity Coefficient of 83.6±6.3% in training and 71.8±10.7% in testing.

Subfields: Computer Vision and Pattern Recognition, Health Information Management, Radiology, Nuclear Medicine and Imaging

How researchers see this paper

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

In Computer Vision and Pattern Recognition

In Health Information Management

In Radiology, Nuclear Medicine and Imaging

Citations in the corpus

ChestX-Ray8: Hospital-Scale Chest X-Ray Database and Benchmarks on Weakly-Supervised Classification and Localization of Common Thorax Diseases (2017, background): “There have been recent efforts on creating openly available annotated medical image databases [48, 50, 36, 35] with the studied patient numbers ranging from a few hundreds to two thousands.”
Interactive Medical Image Segmentation Using Deep Learning With Image-Specific Fine Tuning (2017, neutral): “For 3D volumes, patch-based CNNs have been proposed for segmentation of the brain tumor [7] and pancreas [14], and more powerful end-to-end 3D CNNs include V-Net [15], HighRes3DNet [8], and 3D deeply supervised network [16].”
Automatic Liver and Lesion Segmentation in CT Using Cascaded Fully Convolutional Neural Networks and 3D Conditional Random Fields (2016, neutral): “[20,8,23,21,19,12].”
Automatic Detection of Cerebral Microbleeds From MR Images via 3D Convolutional Neural Networks (2016, methodology): “One straightforward way is to employ conventional 2D CNNs based on a single slice and process the slices sequentially [25]–[27].”
BrainNetCNN: Convolutional neural networks for brain networks; towards predicting neurodevelopment (2017, background): “Artificial Neural Networks (ANNs),(2) specifically Convolutional Neural Networks (CNNs), have had much success lately in performing prediction tasks on medical image data (Cireşan et al., 2012, 2013; Roth et al., 2015).”
Performance of an Artificial Multi-observer Deep Neural Network for Fully Automated Segmentation of Polycystic Kidneys (2017, methodology): “For instance, deep learning techniques have been used in organ [3] and tumor segmentation tasks [4], as well as tissue and tumor classification [5, 6].”
Deep Learning for Multi-Task Medical Image Segmentation in Multiple Modalities (2016, methodology): “Furthermore, the implementation of fully connected layers as 1 × 1 convolution layers and the omission of downsampling layers allowed fast processing of whole images compared with more time-consuming patch-based scanning [11,12,18,10].”
A Fixed-Point Model for Pancreas Segmentation in Abdominal CT Scans (2016, methodology): “We evaluate our approach on the NIH pancreas segmentation dataset [9].”
Regional Homogeneity: Towards Learning Transferable Universal Adversarial Perturbations Against Defenses (2019, methodology): “, adversarial training with feature denoising [62], see Figure 1(b)), generating different types of adversarial examples (image-dependent perturbations or universal ones [50], see Figure 1(c)), or tested on different data domains (CT scans from NIH pancreas segmentation dataset [46], see Figure 1(d)”
Recurrent residual U-Net for medical image segmentation (2019, neutral): “40, and VoxResNet was proposed in 2016 where a deep voxel-wise residual network was used for brain segmentation.”
Medical Image Computing and Computer-Assisted Intervention − MICCAI 2017: 20th International Conference, Quebec City, QC, Canada, September 11-13, 2017, Proceedings, Part II (2017, neutral): “The precision obtained here is particularly significant if we consider the relatively low agreement rates between doctors in determining treatments for LSS [12,13].”
Automatic Liver and Tumor Segmentation of CT and MRI Volumes using Cascaded Fully Convolutional Neural Networks (2017, neutral): “[18, 19, 20, 21, 22, 23, 24, 25].”