Dense biased networks with deep priori anatomy and hard region adaptation: Semi-supervised learning for fine renal artery segmentation

He, Yuting; Yang, Guanyu; Yang, Jian; Chen, Yang; Kong, Yaguang; Wu, Jiasong; Tang, Lijun; Zhu, Xiaomei; Dillenseger, Jean-Louis; Shao, Pengfei; Zhang, Shaobo; Hu, Shu; Coatrieux, Jean-Louis; Li, Shuo

doi:10.1016/j.media.2020.101722

Cited by 60 publications

(37 citation statements)

References 53 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We follow DenseBiasNet structure (He et al, 2020) as our segmentation network for our IRS segmentation and take continuous convolutions, pooling, ac- tivations, and full connections as our meta perceiver for the perception of discriminative compositions in features.…”

Section: Details Of the Networkmentioning

confidence: 99%

“…As shown in Fig. 7 (a), we follow DenseBiasNet (He et al, 2020) as our segmentation network for IRS segmentation, it fuses multi-receptive fields and multi resolution features for the adaptation of scale changes and has achieved success in renal artery segmentation. We take seven resolution stages and each stage has two 3×3×3 convolutional layers followed by a group normalization (Wu and He, 2018)(GN) and a ReLU activation.…”

Section: Densebiasnet For Irs Segmentationmentioning

confidence: 99%

“…In the first stage, the initial window width w 0 and center c 0 are 256 and 256, thus 32 sub-windows on our grayscale window grid are searched. The optimization setting of the corresponding segmentation networks (DenseBi-asNet (He et al, 2020)) trained in these sub-windows is the Adam with the batch size of 1, learning rate of 1 × 10 −4 and iterations of 1.6 × 10 5 . The DSC of their results is ranked for the four best sub-windows.…”

Section: Experiments Configurationsmentioning

confidence: 99%

“…Comparison settings. To illustrate the superiority of our framework, we compare the proposed MGANet with V-Net (Milletari et al, 2016), 3D U-Net (Ç içek et al, 2016), Res-U-Net (Li et al, 2018), Kid-Net (Taha et al, 2018), Dense-BiasNet (He et al, 2020), and an ensemble model, VFN (Xia et al, 2018) on CTA images. To further study the nature of our GIS strategy, we analyze the performance of the segmentation network in different sub-windows, and the variation of their ensemble performance with the number of integrated learners changes.…”

Section: Experiments Configurationsmentioning

confidence: 99%

“…Evaluation metric. The segmentation performance of each method is evaluated in two aspects (Taha and Hanbury, 2015) following He et al (2020). 1) The area-based metric: we take the Dice coefficient (DSC) to evaluate the area-based overlap index.…”

Section: Experiments Configurationsmentioning

confidence: 99%

See 4 more Smart Citations

Meta grayscale adaptive network for 3D integrated renal structures segmentation

Yang

et al. 2021

Medical Image Analysis

Self Cite

View full text Add to dashboard Cite

Three-dimensional (3D) integrated renal structures (IRS) segmentation targets segmenting the kidneys, renal tumors, arteries, and veins in one inference. Clinicians will benefit from the 3D IRS visual model for accurate preoperative planning and intraoperative guidance of laparoscopic partial nephrectomy (LPN). However, no success has been reported in 3D IRS segmentation due to the inherent challenges in grayscale distribution: low contrast caused by the narrow task-dependent distribution range of regions of interest (ROIs), and the networks representation preferences caused by the distribution variation inter-images. In this paper, we propose the Meta Greyscale Adaptive Network (MGANet), the first deep learning framework to simultaneously segment the kidney, renal tumors, arteries and veins on CTA images in one inference. It makes innovations in two collaborate aspects: 1) The Grayscale Interest Search (GIS) adaptively focuses segmentation networks on task-dependent grayscale distributions via scaling the window width and center with two cross-correlated coefficients for the first time, thus learning the fine-grained representation for fine segmentation. 2) The Meta Grayscale Adaptive (MGA) learning makes an image-level meta-learning strategy. It represents diverse robust features from multiple distributions, perceives the distribution characteristic, and generates the model parameters to fuse features dynamically according to image's distribution, thus adapting the grayscale distribution variation. This study enrolls 123 patients and the average Dice coefficients of the renal structures are up to 87.9%. Fine selection of the task-dependent grayscale distribution ranges and personalized fusion of multiple representations on different distributions will lead to better 3D IRS segmentation quality. Extensive experiments with promising results on renal structures reveal powerful segmentation accuracy and great clinical significance in renal cancer treatment.

show abstract

Section: Details Of the Networkmentioning

confidence: 99%

Section: Densebiasnet For Irs Segmentationmentioning

confidence: 99%

Section: Experiments Configurationsmentioning

confidence: 99%

Section: Experiments Configurationsmentioning

confidence: 99%

Section: Experiments Configurationsmentioning

confidence: 99%

See 3 more Smart Citations

Meta grayscale adaptive network for 3D integrated renal structures segmentation

Yang

et al. 2021

Medical Image Analysis

Self Cite

View full text Add to dashboard Cite

show abstract

Boundary-Aware Network for Kidney Tumor Segmentation

Zhang

Xia

2020

Machine Learning in Medical Imaging

View full text Add to dashboard Cite

Kidney structures segmentation is a crucial yet challenging task in the computer-aided diagnosis of surgery-based renal cancer. Although numerous deep learning models have achieved remarkable success in many medical image segmentation tasks, accurate segmentation of kidney structures on computed tomography angiography (CTA) images remains challenging, due to the variable sizes of kidney tumors and the ambiguous boundaries between kidney structures and their surroundings. In this paper, we propose a boundary-aware network (BA-Net) to segment kidneys, kidney tumors, arteries, and veins on CTA scans. This model contains a shared encoder, a boundary decoder, and a segmentation decoder. The multi-scale deep supervision strategy is adopted on both decoders, which can alleviate the issues caused by variable tumor sizes. The boundary probability maps produced by the boundary decoder at each scale are used as attention to enhance the segmentation feature maps. We evaluated the BA-Net on the Kidney PArsing (KiPA) Challenge dataset and achieved an average Dice score of 89.65% for kidney structure segmentation on CTA scans using 4-fold cross-validation. The results demonstrate the effectiveness of the BA-Net.

show abstract

EnMcGAN: Adversarial Ensemble Learning for 3D Complete Renal Structures Segmentation

et al. 2021

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Dense biased networks with deep priori anatomy and hard region adaptation: Semi-supervised learning for fine renal artery segmentation

Cited by 60 publications

References 53 publications

Meta grayscale adaptive network for 3D integrated renal structures segmentation

Meta grayscale adaptive network for 3D integrated renal structures segmentation

Boundary-Aware Network for Kidney Tumor Segmentation

EnMcGAN: Adversarial Ensemble Learning for 3D Complete Renal Structures Segmentation

Contact Info

Product

Resources

About