Deep learning based classification of breast tumors with shear-wave elastography

Zhang, Qi; Xiao, Yang; Dai, Wei; Suo, Jingfeng; Wang, Congzhi; Shi, Jun; Zheng, Hairong

doi:10.1016/j.ultras.2016.08.004

Cited by 201 publications

(109 citation statements)

References 38 publications

Supporting

Mentioning

105

Contrasting

Unclassified

Order By: Relevance

“…Transfer learning methods were employed for breast mass classification and segmentation in several studies . Additionally, deep learning was used to detect breast lesions and differentiate breast masses with shear‐wave elastography . The better‐performing pretrained deep learning models have been developed using RGB color images .…”

Section: Introductionmentioning

confidence: 99%

“…[19][20][21][22][23][24][25] Additionally, deep learning was used to detect breast lesions 26 and differentiate breast masses with shear-wave elastography. 27 The betterperforming pretrained deep learning models have been developed using RGB color images. [28][29][30][31] However, medical images, including US images, are commonly grayscale, which raises question about how to efficiently utilize the discriminative power of a pretrained model.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Breast mass classification in sonography with transfer learning using a deep convolutional neural network and color conversion

et al. 2019

View full text Add to dashboard Cite

Purpose We propose a deep learning‐based approach to breast mass classification in sonography and compare it with the assessment of four experienced radiologists employing breast imaging reporting and data system 4th edition lexicon and assessment protocol. Methods Several transfer learning techniques are employed to develop classifiers based on a set of 882 ultrasound images of breast masses. Additionally, we introduce the concept of a matching layer. The aim of this layer is to rescale pixel intensities of the grayscale ultrasound images and convert those images to red, green, blue (RGB) to more efficiently utilize the discriminative power of the convolutional neural network pretrained on the ImageNet dataset. We present how this conversion can be determined during fine‐tuning using back‐propagation. Next, we compare the performance of the transfer learning techniques with and without the color conversion. To show the usefulness of our approach, we additionally evaluate it using two publicly available datasets. Results Color conversion increased the areas under the receiver operating curve for each transfer learning method. For the better‐performing approach utilizing the fine‐tuning and the matching layer, the area under the curve was equal to 0.936 on a test set of 150 cases. The areas under the curves for the radiologists reading the same set of cases ranged from 0.806 to 0.882. In the case of the two separate datasets, utilizing the proposed approach we achieved areas under the curve of around 0.890. Conclusions The concept of the matching layer is generalizable and can be used to improve the overall performance of the transfer learning techniques using deep convolutional neural networks. When fully developed as a clinical tool, the methods proposed in this paper have the potential to help radiologists with breast mass classification in ultrasound.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Breast mass classification in sonography with transfer learning using a deep convolutional neural network and color conversion

et al. 2019

View full text Add to dashboard Cite

show abstract

“…The velocity of laterally propagated shear waves (lateral waves) is measured. SWE is useful for a diagnosis of breast tumor[1], thyriod tumor[2], muscle stiffness[3] as well as liver stiffness. SWE resembles acoustic radiation force impulse[4], but it is new another technology.…”

Section: Introductionmentioning

confidence: 99%

Shear wave elastography in hepatitis C patients before and after antiviral therapy

Suda¹,

Ohashi²,

Masaoka³

et al. 2017

WJH

View full text Add to dashboard Cite

AIMTo investigate shear wave (SW) propagation velocity in patients with untreated hepatitis C and patients with sustained virological response (SVR).METHODSA total of 136 hepatitis C patients [85 patients who had not received antiviral therapy (naïve group) and 51 patients who had received antiviral therapy and subsequently achieved SVR of at least 24 wk (SVR group)] and 58 healthy volunteers and outpatients without liver disease (control group) underwent evaluation of liver stiffness by SW elastography (SWE). Various parameters were evaluated in the chronic hepatitis C patients at the time of SWE.RESULTSSW propagation velocity (Vs) was 1.23 ± 0.14 m/s in the control group, 1.56 ± 0.32 m/s in the SVR group, and 1.69 ± 0.31 m/s in the naïve group. Significant differences were seen between the control group and the SVR group (P = 0.0000) and between the SVR group and the naïve group (P = 0.01417). All four fibrosis markers were higher in the naïve group than in the SVR group. In the naïve group, Vs was positively correlated with alanine aminotransferase (ALT) (r = 0.5372), α feto protein (AFP) (r = 0.4389), type IV collagen (r = 0.5883), procollagen III peptide (P-III-P) (r = 0.4140), hyaluronic acid (r = 0.4551), and Mac-2 binding protein glycosylation isomer (M2BPGi) (r = 0.6092) and negatively correlated with albumin (r = -0.4289), platelets (r = -0.5372), and prothrombin activity (r = -0.5235). On multiple regression analysis, Vs was the most strongly correlated with ALT (standard partial regression std β = 0.4039, P = 0.00000). In the SVR group, Vs was positively correlated with AFP (r = 0.6977), type IV collagen (r = 0.5228), P-III-P (r = 0.5812), hyaluronic acid (r = 0.5189), and M2BPGi (r = 0.6251) and negatively correlated with albumin (r = -0.4283), platelets (r = -0.4842), and prothrombin activity (r = -0.4771). On multiple regression analysis, Vs was strongly correlated with AFP (standard partial regression std β = 0.5953, P = 0.00000) and M2BPGi (standard partial regression std β= 0.2969, P = 0.03363).CONCLUSIONIn hepatitis C patients, liver stiffness is higher in treatment-naïve patients than in those showing SVR. SWE may be a predictor of hepatocarcinogenesis in SVR patients.

show abstract

“…Many studies have shown that tumor detection using CNNs is useful in breast cancer screening. Most of the results are from mammography [21,22,23,24], but studies using tomosynthesis [25,26], ultrasound [27], and MRI [28] have also been published.…”

Section: Cancer Screening In Mammographymentioning

confidence: 99%

Deep Learning for Cancer Screening in Medical Imaging

Jeong

2017

Hanyang Med Rev

View full text Add to dashboard Cite

show abstract

Deep learning based classification of breast tumors with shear-wave elastography

Cited by 201 publications

References 38 publications

Breast mass classification in sonography with transfer learning using a deep convolutional neural network and color conversion

Breast mass classification in sonography with transfer learning using a deep convolutional neural network and color conversion

Shear wave elastography in hepatitis C patients before and after antiviral therapy

Deep Learning for Cancer Screening in Medical Imaging

Contact Info

Product

Resources

About