Automated fibroglandular tissue segmentation in breast MRI using generative adversarial networks.

08:00 EDT 10th March 2020 | BioPortfolio

Summary of "Automated fibroglandular tissue segmentation in breast MRI using generative adversarial networks."

Fibroglandular tissue (FGT) segmentation is a crucial step for quantitative analysis of background parenchymal enhancement (BPE) in MRI, which is useful for breast cancer risk assessment. In this study, we develop an automated deep learning method based on a generative adversarial network (GAN) to identify the FGT region in MRI volumes and evaluate its impact on a specific clinical application. The GAN consists of an improved U-Net as a generator to generate FGT candidate areas and a patch deep convolutional neural network (DCNN) as discriminator to evaluate the authenticity of the synthetic FGT region. The proposed method has two improvements compared to the classical U-Net: 1) the improved U-Net is designed to extract more features of the FGT region for a more accurate description of the FGT region; 2) a patch DCNN is designed for discriminating the authenticity of the FGT region generated by the improved U-Net, which makes the segmentation result more stable and accurate. A dataset of 100 three-dimensional bilateral breast MRI scans from 100 patients (aged 22-78 years) was used in this study with the Institutional Review Board (IRB) approval. Three-dimensional hand segmented FGT areas for all breasts were provided as a reference standard. Five-fold cross-validation was used in training and testing of the models. The Dice similarity coefficient (DSC) and Jaccard index (JI) values were evaluated to measure the segmentation accuracy. Previous method using classical U-Net was used as a baseline in this study. In the five partitions of the cross-validation set, the GAN achieved DSC and JI values of 87.0±7.0% and 77.6±10.1% respectively, while the corresponding values obtained through by the baseline method were 81.1±8.7% and 69.0±11.3% respectively. The proposed method is significantly superior to the previous method using U-Net. The FGT segmentation impacted the BPE quantification application in the following manner: the correlation coefficients between quantified BPE value and BI-RADS BPE categories provided by the radiologist were 0.46±0.15 (best: 0.63) based on GAN segmented FGT areas, while the corresponding correlation coefficients were 0.41±0.16 (best: 0.60) based on baseline U-Net segmented FGT areas. BPE can be quantified better using the FGT areas segmented by the proposed GAN model than using the FGT areas segmented by the baseline U-Net.


Journal Details

This article was published in the following journal.

Name: Physics in medicine and biology
ISSN: 1361-6560


DeepDyve research library

PubMed Articles [10215 Associated PubMed Articles listed on BioPortfolio]

Semi-supervised segmentation of lesion from breast ultrasound images with attentional generative adversarial network.

Automatic segmentation of breast lesion from ultrasound images is a crucial module for the computer aided diagnostic systems in clinical practice. Large-scale breast ultrasound (BUS) images remain una...

Automated volumetric breast density measures: differential change between breasts in women with and without breast cancer.

Given that breast cancer and normal dense fibroglandular tissue have similar radiographic attenuation, we examine whether automated volumetric density measures identify a differential change between b...

Generative adversarial network in medical imaging: A review.

Generative adversarial networks have gained a lot of attention in the computer vision community due to their capability of data generation without explicitly modelling the probability density function...

Knowledge-based automated planning with three-dimensional generative adversarial networks.

To develop a knowledge-based automated planning pipeline that generates treatment plans without feature engineering, using deep neural network architectures for predicting 3D doses.

Joint optic disc and cup segmentation using semi-supervised conditional GANs.

Glaucoma is a chronic and widespread eye disease threatening humans' irreversible vision loss. The cup-to-disc ratio (CDR), one of the most important measurements used for glaucoma screening and diagn...

Clinical Trials [9917 Associated Clinical Trials listed on BioPortfolio]

Fully Automated Pipeline for the Detection and Segmentation of Non-Small Cell Lung Cancer (NSCLC) on CT Images

Accurate segmentation of lung tumor is essential for treatment planning, as well as for monitoring response to therapy. It is well-known that segmentation of the lung tumour by different r...

Gradient-Echo Spectroscopic Imaging Study of Saturated Fat and Breast Cancer

The role of fat in breast cancer development and growth has been studied extensively using body mass index, a measure of whole body fatness, and dietary fat intake in a number of epidemiol...

Case Collection Study to Determine the Accuracy, Call Back and Cancer Detection Rates of QT Ultrasound in Breast Imaging

Brief Summary: The QT Ultrasound system is an automated scanner which transmits pulsed ultrasound plane waves through the breast, as well as collects reflected ultrasound output. As the pa...

Earlier Breast Cancer Detection Using Automated Whole Breast Ultrasound With Mammography, Including Cost Comparisons

The purpose of this study is to determine whether addition of automated whole breast ultrasound to the usual screening mammography in a population of asymptomatic women with mammographical...

Building Sentences With Preschoolers Who Use AAC

The focus of this investigation is to compare the effectiveness of the AAC Generative Language Intervention approach to an AAC Standard of Care condition on preschool sentence productions....

Medical and Biotech [MESH] Definitions

A common and benign breast disease characterized by varying degree of fibrocystic changes in the breast tissue. There are three major patterns of morphological changes, including FIBROSIS, formation of CYSTS, and proliferation of glandular tissue (adenosis). The fibrocystic breast has a dense irregular, lumpy, bumpy consistency.

Measurement of relative composition of different BREAST tissue types often determined from MAMMOGRAPHY; ULTRASONOGRAPHY; or MRI.

Excision of breast tissue with preservation of overlying skin, nipple, and areola so that breast form may be reconstructed.

A fluid-filled closed cavity or sac that is lined by an EPITHELIUM and found in the BREAST. It may appear as a single large cyst in one breast, multifocal, or bilateral in FIBROCYSTIC BREAST DISEASE.

Neoplasms, usually carcinoma, located within the center of an organ or within small lobes, and in the case of the breast, intraductally. The emphasis of the name is on the location of the neoplastic tissue rather than on its histological type. Most cancers of this type are located in the breast.

Quick Search

DeepDyve research library

Relevant Topics

Breast Cancer
Track and monitor developments in breast cancer research and commercial development.  Follow the tabs above to read the latest global news, research, clinical trials on breast cancer and follow companies active in the development of breast cancer tr...

Radiology is the branch of medicine that studies imaging of the body; X-ray (basic, angiography, barium swallows), ultrasound, MRI, CT and PET. These imaging techniques can be used to diagnose, but also to treat a range of conditions, by allowing visuali...

Benign Prostatic Hyperplasia (BPH)
An enlarged prostate (benign prostatic enlargement (BPE) or benign prostatic hyperplasia (BPH)) is common in men after the age of about 50. Having an enlarged prostate does not mean you have cancer. In some cases, an enlarged prostate can cause the ...

Searches Linking to this Article