Detection and classification the breast tumors using mask R-CNN on sonograms

Chiao, Jui-Ying; Chen, Kuan-Yung; Liao, Ken Ying-Kai; Hsieh, Po-Hsin; Zhang, Geoffrey; Huang, Tzung-Chi

doi:10.1097/md.0000000000015200

Cited by 151 publications

(82 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…This paper improves the Mask R-CNN [12] to extract image features on webpages. The Mask R-CNN ( Figure 1) is a combination of two classical target recognition algorithms: Fast R-CNN and fully convolutional network (FCN).…”

Section: Image Feature Extraction Based On Deep Learningmentioning

confidence: 99%

See 1 more Smart Citation

A Malicious Webpage Detection Algorithm Based on Image Semantics

Li¹,

Li²,

Liu³

et al. 2020

View full text Add to dashboard Cite

In the era of the Internet, malicious attacks have put user information at risk. Many malicious webpages use images as the carrier of malicious codes. If extracted accurately, the features of these images will help to improve the detection of malicious webpages. This paper aims to develop an accurate malicious webpage detection method based on the features of the said images. Since static images contain a small amount of information, semantic segmentation was performed to predict the semantics of the target attitude. Then, the final semantics of target the image were derived by the backpropagation neural network (BPNN). After that, the image semantics were fused with the other features of the malicious webpage, and sent to the classifier for recognition. Finally, the proposed algorithm was tested on an actual dataset, in comparison with other malicious webpage detection methods. The results show that our algorithm can accurately detect malicious webpages, thanks to the introduction of image semantic features.

show abstract

Section: Image Feature Extraction Based On Deep Learningmentioning

confidence: 99%

“…The URL contains the character @. 12 The registration is less than 4 months. 13 The webpage ranks below 5 .…”

mentioning

confidence: 99%

A Malicious Webpage Detection Algorithm Based on Image Semantics

Li¹,

Li²,

Liu³

et al. 2020

View full text Add to dashboard Cite

show abstract

“…Medical imaging techniques, such as magnetic resonance imaging (MRI), mammography, and ultrasound, are commonly used in breast disorders because they are noninvasive and provide information about the overall anatomy of the breast tumor. MRI scans are super-sensitive to soft tissue lesions [4]. However, they cost a lot of time and money than other available tests and contain a relatively high number of contraindications.…”

Section: (Iii) Full Text Introductionmentioning

confidence: 99%

“…Mammography is super-sensitive to the detection of calcification, but its sensitivity decreases by 30% when scanning dense breast tissue [4]. The breast density of Asian women is generally higher than that of Western women, which slightly restricts the application of mammography in patients from Asian countries [4].…”

Section: (Iii) Full Text Introductionmentioning

confidence: 99%

Image-Based Ultrasomics Classifiers: A Potential Method to Noninvasively Predict the Proliferation-Related Biomarker Ki67

Yuan¹,

Xu²,

Fang³

et al. 2020

Preprint

View full text Add to dashboard Cite

Objectives: The proliferation-related biomarker Ki67 is excellent in predicting the prognosis of breast cancer. 3D ultrasound imaging can provide information about coronal section images that help diagnose breast cancer. Our study aims to develop an ultrasomics method that extracts information related to Ki67 by analyzing the maximum transverse/sagittal/coronal section images of breast cancer mass. Methods: Our study retrospectively collected data on patients who had finished 3D ultrasound examinations and were pathologically diagnosed with breast cancer. Images met the criteria were segmented, and then regions of interest (ROIs) were outlined for extracting ultrasomics features, such as statistical, morphological, texture, filter, and wavelet features.Results: The least absolute shrinkage and selection operator (LASSO) regression model selected 16 features that were closely related to the Ki67. The classification results of sensitivity, specificity, accuracy, and area under curve (AUC) of the transverse-sectional images were 0.6451, 0.8064, 0.7258, and 0.8065 (95% CI, 0.6915-0.9214), respectively; for sagittal-sectional images were 0.5806, 0.7741, 0.6774, and 0.6660 (95% CI, 0.5283-0.8037) respectively; for coronal-sectional images were 0.5806, 0.6774, 0.6290, and 0.7159 (95% CI, 0.5847-0.8471) respectively, and for a combination of three-section images were 0.7667, 0.7500, 0.7580, and 0.8510 (95% CI, 0.7537 -0.9483).Conclusions: The model classifier based on the transverse section images performed better than that based on the sagittal/coronal section images. The model classifier based on a combination of three-section images had a better outcome than that used only single section images. Image-based ultrasomics classifiers can noninvasively predict the Ki67 of breast cancer.

show abstract

“…Deep learning, a type of arti cial neural network, can automatically process larger and more complex data sets using technologies such as convolutional neural networks. It has been applied to the diagnosis of thyroid nodules [28], as well as the detection of breast tumors [29,30] and skin cancers [31], and might be capable of addressing current limitations in ML for MI analysis.…”

Section: Introductionmentioning

confidence: 99%

Deep Polar Residual Network for Identifying Myocardial Infarction in Mce Imaging

Shen

Xue

Guo

et al. 2020

Preprint

View full text Add to dashboard Cite

Background: Ultrasonic diagnosis of myocardial infarction (MI) varies markedly among radiologists, and is also time-consuming. In addition, there is a high level of inter-observer variation in diagnosis of MI. The purpose of this study is to develop a deep learning-based model that automatically recognizes MI in myocardial contrast echocardiography(MCE) images, and to compare the model performance against radiologists with varying experience in echocardiography.Methods: In this study, a polar residual network (PResNet) model was employed to recognize MI segments in MCE images of Sprague-Dawley (SD) rats. All SD rats were randomly divided into the MI model group (n=15) and the sham group (n=10). Rats in MI model group were infarcted via left anterior descending coronary artery ligation. Ultimately, MCE was performed on SD rats to observe their myocardium. Left ventricle (LV) short-axis views of MCE at the basal and midpapillary muscle levels for each SD rat were divided into 6 segments separately. The PResNet model and six radiologists each diagnosed 276 MCE image segments. Masson’s trichrome staining was used as the gold standard for MI presence or absence. The PResNet model and radiologist diagnostic performance were evaluated using sensitivity, specificity, and area under the receiver operating characteristic curve (AUC). Furthermore, analysis duration per rat and interobserver variability were also compared. Results: Masson’s trichrome stain revealed 30 infarcted segments out of the 276 total segments. The PResNet model achieved an AUC of 0.89 (0.85-0.93), which was comparable to that of senior radiologists (0.89; 0.85-0.93). Relative to senior radiologists, the PResNet model presented slightly lower sensitivity but significantly higher specificity. The mean sensitivity and specificity for the junior radiologists were significantly lower than both the PResNet model and the senior radiologists. In terms of analysis duration, the PResNet model (148-169sec) was markedly faster than both junior (314-366sec) and senior (193-235sec). Moreover, the intrasample agreement for the PResNet model (k=0.96) was better than the interobserver agreement for the junior (k=0.81) and senior radiologists (k=0.86). Conclusion: The PResNet model not only improved diagnostic performance and observer consensus but also improve the diagnose speed. It is a promising tool for identifying MI.

show abstract

Detection and classification the breast tumors using mask R-CNN on sonograms

Cited by 151 publications

References 17 publications

A Malicious Webpage Detection Algorithm Based on Image Semantics

A Malicious Webpage Detection Algorithm Based on Image Semantics

Image-Based Ultrasomics Classifiers: A Potential Method to Noninvasively Predict the Proliferation-Related Biomarker Ki67

Deep Polar Residual Network for Identifying Myocardial Infarction in Mce Imaging

Contact Info

Product

Resources

About