Posts Tagged ‘3D U-Net’
Complete List of Challenge Submissions
An attempt at beating the 3D U-Net
Isensee, Fabian
Maier-Hein, Klaus H.
The U-Net is arguably the most successful segmentation architecture in the medical domain. Here we apply a 3D U-Net to the 2019 Kidney and Kidney Tumor Segmentation Challenge and attempt to improve upon it by augmenting it with residual and pre-activation residual blocks. Cross-validation results on the training cases suggest only very minor, barely measurable improvements. Due to marginally higher dice scores, the residual 3D U-Net is chosen for test set prediction. With a Composite Dice score of 91.23 on the test set, our method outperformed all 105 competing teams and won the KiTS2019 challenge by a small margin.
3D Unet-based Kidney and Kidney Tumer Segmentation with Attentive Feature Learning
Tao, Qingyi
Wu, Zhonghua
Cheong, Isaac
Yang, Jingyi
Ge, Zongyuan
Lin, Guosheng
Cai, Jianfei
To study the kidney diseases and kidney tumor from Computed Tomography(CT) imaging data, it is helpful to segment the region of interest through computer aided auto-segmentation tool. In the KiTs 2019 challenge [1], we are provided 3D volumetric CT data to train a model for kidney and kidney tumor segmentation. We introduce an improved deep 3D Unet by enriching the feature representation in CT images using an attention module. We achieve 1.5% improvement in the segmentation accuracy when evaluated on the validation set.
Kidney Tumor Segmentation Using 3D U-nets for Masking and Labeling
Ito, Yuichi
Nishimoto, Takuya
Chou, Weizhong
Ishiguro, Toshitaka
Mori, Kensaku
Kojo, Kosuke
Kojima, Takahiro
Nakagawa, Tsutomu
Kakeya, Hideki
In this paper we report on the algorithm we applied to KiTS 19 Challenge. We used 3D U-nets for masking and labeling separately. We mainly focused on data augmentation and grouping of training data to improve the result of segmentation.
Using Two-stage Network to Segment Kidneys and Kidney Tumors
Chen, Pan
Xu, Chenghai
He, Jie
Sun, Chengwei
Ma, Yingying
Sun, Fenglong
There are many new cases of kidney cancer each year, and surgery is the most common treatment. To assist doctors in surgical planning, an accurate and automatic kidney and tumor segmentation method is helpful in the clinical practice. In this paper, we propose a deep learning framework for the segmentation of kidneys and tumors in abdominal CT images. The key idea is using a two-stage strategy. First, for each case, we use a 3d U-shape convolution network to get the localization of each kidney. Then using next 3d U-shape convolution network we obtain the precise segmentation results of each kidney. Finally, merge the results to obtain the complete segmentation. Also, we try some tricks to improve the performance.
Attention Guided 3D U-Net for KiTS19
Zhong, Zhusi
Zhang, Zhenxi
Jiao, Zhicheng
We use a two-stage 3d U-Net model to predict the multi channels segmentations from coarse to fine. The second stage is guided by the predictions from the first stage. 1 Method We proposed a two stages method to segment CT image from coarse to fine. The two stages are trained with different learning scope and are assigned with different learning missions. 1.1 Stage 1 – Coarse stage Data preprocess. Firstly, we downscale the training data to a normal shape, in order to make sure the model can take a whole image at once. All the images and segmentations are downscale to 128*128*32 (height*width*depth). The segmentation files are transformed to 3-channels arrays, in which the channels-wise pixel values represent kidneys, tumors and the background (without kidneys and tumors) in order. Training. We train the standard 3D U-Net follow with a softmax layer. While training, we apply some data augmentation to the training data, including normalize, random contrast, random flip and random rotate. We input all the 210 cases training data and train the model to regress the multi-channel segmentations. We apply with pytorch, and the learning rate is 0.1 which divide 0.1 in 300000 epochs and 500000 epochs. We use the Binary Cross Entropy Loss as loss function. Predicting. The 90 cases testing images are preprocessed the same with the training images then input to the trained model. The channel-wise predictions are scaled back the original shape. We take the first 2 channels of the predictions, represent as the segmentation of kidneys and tumors, then package as the .nii.gz files.