A novel convolutional neural network for kidney ultrasound images segmentation

Chen, Gongping; Yin, Jingjing; Dai, Yu; Zhang, Jianxun; Yin, Xin; Cui, Liang

doi:10.1016/j.cmpb.2022.106712

Cited by 27 publications

(7 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For the comparison of breast ultrasound, we also selected five specific segmentation methods, AE U-Net [17], STAN [33], RDAU-Net [34], MADU-net [10], SKU-net [21]. Similarly, we used three specific methods SDFNet [35], MBANet [36], and MSDSNet [37] for kidney ultrasound. According to the opensource codes, we retrain these comparison networks on the same datasets as our method.…”

Section: Resultsmentioning

confidence: 99%

Asymmetric U-shaped network with hybrid attention mechanism for kidney ultrasound images segmentation

Chen

Zhao²,

Dai³

et al. 2023

Expert Systems with Applications

Self Cite

View full text Add to dashboard Cite

Section: Resultsmentioning

confidence: 99%

Asymmetric U-shaped network with hybrid attention mechanism for kidney ultrasound images segmentation

Chen

Zhao²,

Dai³

et al. 2023

Expert Systems with Applications

Self Cite

View full text Add to dashboard Cite

“…The deep learning model trained by labeled images can directly process the raw data and standardized the segmented region of interest through the neural network, therefore additional artificial error could be avoided. The use of CNN-based models is highly effective in several imaging modalities, however, it is still challenging for well-known speckle noises, serious cascade, uneven intensity distribution and blurred boundaries in gray-scale ultrasound images (30,31).…”

Section: Discussionmentioning

confidence: 99%

CNN-based automatic segmentations and radiomics feature reliability on contrast-enhanced ultrasound images for renal tumors

Yang,

Chen,

Liang

et al. 2023

Front. Oncol.

View full text Add to dashboard Cite

ObjectiveTo investigate the feasibility and efficiency of automatic segmentation of contrast-enhanced ultrasound (CEUS) images in renal tumors by convolutional neural network (CNN) based models and their further application in radiomic analysis.Materials and methodsFrom 94 pathologically confirmed renal tumor cases, 3355 CEUS images were extracted and randomly divided into training set (3020 images) and test set (335 images). According to the histological subtypes of renal cell carcinoma, the test set was further split into clear cell renal cell carcinoma (ccRCC) set (225 images), renal angiomyolipoma (AML) set (77 images) and set of other subtypes (33 images). Manual segmentation was the gold standard and serves as ground truth. Seven CNN-based models including DeepLabV3+, UNet, UNet++, UNet3+, SegNet, MultilResUNet and Attention UNet were used for automatic segmentation. Python 3.7.0 and Pyradiomics package 3.0.1 were used for radiomic feature extraction. Performance of all approaches was evaluated by the metrics of mean intersection over union (mIOU), dice similarity coefficient (DSC), precision, and recall. Reliability and reproducibility of radiomics features were evaluated by the Pearson coefficient and the intraclass correlation coefficient (ICC).ResultsAll seven CNN-based models achieved good performance with the mIOU, DSC, precision and recall ranging between 81.97%-93.04%, 78.67%-92.70%, 93.92%-97.56%, and 85.29%-95.17%, respectively. The average Pearson coefficients ranged from 0.81 to 0.95, and the average ICCs ranged from 0.77 to 0.92. The UNet++ model showed the best performance with the mIOU, DSC, precision and recall of 93.04%, 92.70%, 97.43% and 95.17%, respectively. For ccRCC, AML and other subtypes, the reliability and reproducibility of radiomic analysis derived from automatically segmented CEUS images were excellent, with the average Pearson coefficients of 0.95, 0.96 and 0.96, and the average ICCs for different subtypes were 0.91, 0.93 and 0.94, respectively.ConclusionThis retrospective single-center study showed that the CNN-based models had good performance on automatic segmentation of CEUS images for renal tumors, especially the UNet++ model. The radiomics features extracted from automatically segmented CEUS images were feasible and reliable, and further validation by multi-center research is necessary.

show abstract

“…This approach could significantly aid the clinical interpretation of kidney ultrasound images. Xie et al [10] developed a technique for kidney segmentation using texture and shape priors to accurately delineate the kidney's boundary. Texture features are extracted using Gabor filters, and a texture model is created to assess similarities in internal and external regions.…”

Section: Related Workmentioning

confidence: 99%

Coarse-to-fine approach: Automatic delineation of kidney ultrasound data

Peng,

Ruan,

et al. 2024

Big Data Min. Anal.

View full text Add to dashboard Cite

We present an automatic kidney segmentation method using ultrasound images. This method employs a coarse-to-fine approach to tackle the challenge of unclear and fuzzy boundaries. Four key innovations are introduced to enhance the segmentation process's accuracy and efficiency. First, an automatic deep fusion training network serves as a coarse segmentation strategy. Second, we propose an explainable mathematical mapping formula to better represent the kidney contour. Third, by utilizing the characteristics of the principal curve, a neural network automatically refines curve shapes, thus reducing model errors. Finally, we employ an intelligent searching polyline segment method for automatic kidney contour segmentation. The results show that our method achieves high accuracy and stability in segmenting kidney ultrasound images. This work's contributions include the deep fusion training network, intelligent searching polyline segment method, and explainable mathematical mapping formula, which are applicable to other medical image segmentation tasks. Additionally, this approach uses a mean-shift clustering model, supplanting standard projection and vertex optimization steps.

show abstract

A novel convolutional neural network for kidney ultrasound images segmentation

Cited by 27 publications

References 27 publications

Asymmetric U-shaped network with hybrid attention mechanism for kidney ultrasound images segmentation

Asymmetric U-shaped network with hybrid attention mechanism for kidney ultrasound images segmentation

CNN-based automatic segmentations and radiomics feature reliability on contrast-enhanced ultrasound images for renal tumors

Coarse-to-fine approach: Automatic delineation of kidney ultrasound data

Contact Info

Product

Resources

About