An Automatic Segmentation of Gland Nuclei in Gastric Cancer Based on Local and Contextual Information

AbstractSegmenting a liver and its peripherals from abdominal computed tomography is a crucial step toward computer aided diagnosis and therapeutic intervention. Despite the recent advances in computing methods, faithfully segmenting the liver has remained a challenging task, due to indefinite boundary, intensity inhomogeneity, and anatomical variations across subjects. In this paper, a semi-automatic segmentation method based on multivariable normal distribution of liver tissues and graph-cut sub-division is presented. Although it is not fully automated, the method minimally involves human interactions. Specifically, it consists of three main stages. Firstly, a subject specific probabilistic model was built from an interior patch, surrounding a seed point specified by the user. Secondly, an iterative assignment of pixel labels was applied to gradually update the probabilistic map of the tissues based on spatio-contextual information. Finally, the graph-cut model was optimized to extract the 3D liver from the image. During post-processing, overly segmented nodal regions due to fuzzy tissue separation were removed, maintaining its correct anatomy by using robust bottleneck detection with adjacent contour constraint. The proposed system was implemented and validated on the MICCAI SLIVER07 dataset. The experimental results were benchmarked against the state-of-the-art methods, based on major clinically relevant metrics. Both visual and numerical assessments reported herein indicated that the proposed system could improve the accuracy and reliability of asymptomatic liver segmentation.

Download Full-text

Predicting Chemotherapeutic Response for Far-advanced Gastric Cancer by Radiomics with Deep Learning Semi-automatic Segmentation

Journal of Cancer ◽

10.7150/jca.46704 ◽

2020 ◽

Vol 11 (24) ◽

pp. 7224-7236

Author(s):

Jing-wen Tan ◽

Lan Wang ◽

Yong Chen ◽

WenQi Xi ◽

Jun Ji ◽

...

Keyword(s):

Gastric Cancer ◽

Deep Learning ◽

Advanced Gastric Cancer ◽

Automatic Segmentation ◽

Chemotherapeutic Response

Download Full-text

Convolutional neural network for automated mass segmentation in mammography

BMC Bioinformatics ◽

10.1186/s12859-020-3521-y ◽

2020 ◽

Vol 21 (S1) ◽

Author(s):

Dina Abdelhafiz ◽

Jinbo Bi ◽

Reda Ammar ◽

Clifford Yang ◽

Sheida Nabavi

Keyword(s):

Neural Network ◽

Convolutional Neural Network ◽

Data Augmentation ◽

Contextual Information ◽

Automatic Segmentation ◽

Ground Truth ◽

Test Accuracy ◽

Convolutional Network ◽

Proposed Model ◽

The Mean

Abstract Background Automatic segmentation and localization of lesions in mammogram (MG) images are challenging even with employing advanced methods such as deep learning (DL) methods. We developed a new model based on the architecture of the semantic segmentation U-Net model to precisely segment mass lesions in MG images. The proposed end-to-end convolutional neural network (CNN) based model extracts contextual information by combining low-level and high-level features. We trained the proposed model using huge publicly available databases, (CBIS-DDSM, BCDR-01, and INbreast), and a private database from the University of Connecticut Health Center (UCHC). Results We compared the performance of the proposed model with those of the state-of-the-art DL models including the fully convolutional network (FCN), SegNet, Dilated-Net, original U-Net, and Faster R-CNN models and the conventional region growing (RG) method. The proposed Vanilla U-Net model outperforms the Faster R-CNN model significantly in terms of the runtime and the Intersection over Union metric (IOU). Training with digitized film-based and fully digitized MG images, the proposed Vanilla U-Net model achieves a mean test accuracy of 92.6%. The proposed model achieves a mean Dice coefficient index (DI) of 0.951 and a mean IOU of 0.909 that show how close the output segments are to the corresponding lesions in the ground truth maps. Data augmentation has been very effective in our experiments resulting in an increase in the mean DI and the mean IOU from 0.922 to 0.951 and 0.856 to 0.909, respectively. Conclusions The proposed Vanilla U-Net based model can be used for precise segmentation of masses in MG images. This is because the segmentation process incorporates more multi-scale spatial context, and captures more local and global context to predict a precise pixel-wise segmentation map of an input full MG image. These detected maps can help radiologists in differentiating benign and malignant lesions depend on the lesion shapes. We show that using transfer learning, introducing augmentation, and modifying the architecture of the original model results in better performance in terms of the mean accuracy, the mean DI, and the mean IOU in detecting mass lesion compared to the other DL and the conventional models.

Download Full-text

Cross Attention Squeeze Excitation Network (CASE-Net) for Whole Body Fetal MRI Segmentation

Sensors ◽

10.3390/s21134490 ◽

2021 ◽

Vol 21 (13) ◽

pp. 4490

Author(s):

Justin Lo ◽

Saiee Nithiyanantham ◽

Jillian Cardinell ◽

Dylan Young ◽

Sherwin Cho ◽

...

Keyword(s):

Clinical Decision Making ◽

Contextual Information ◽

Automatic Segmentation ◽

Fetal Mri ◽

Clinical Decision ◽

Disease Diagnosis ◽

Whole Body ◽

Resonance Imaging ◽

Mri Segmentation ◽

Magnetic Resonance Imaging Mri

Segmentation of the fetus from 2-dimensional (2D) magnetic resonance imaging (MRI) can aid radiologists with clinical decision making for disease diagnosis. Machine learning can facilitate this process of automatic segmentation, making diagnosis more accurate and user independent. We propose a deep learning (DL) framework for 2D fetal MRI segmentation using a Cross Attention Squeeze Excitation Network (CASE-Net) for research and clinical applications. CASE-Net is an end-to-end segmentation architecture with relevant modules that are evidence based. The goal of CASE-Net is to emphasize localization of contextual information that is relevant in biomedical segmentation, by combining attention mechanisms with squeeze-and-excitation (SE) blocks. This is a retrospective study with 34 patients. Our experiments have shown that our proposed CASE-Net achieved the highest segmentation Dice score of 87.36%, outperforming other competitive segmentation architectures.

Download Full-text

Convolutional neural network for automated mass segmentation in mammography

10.1101/2020.12.01.406975 ◽

2020 ◽

Author(s):

Dina Abdelhafiz ◽

Jinbo Bi ◽

Reda Ammar ◽

Clifford Yang ◽

Sheida Nabavi

Keyword(s):

Neural Network ◽

Convolutional Neural Network ◽

Data Augmentation ◽

Contextual Information ◽

Automatic Segmentation ◽

Ground Truth ◽

Test Accuracy ◽

Convolutional Network ◽

Proposed Model ◽

The Mean

AbstractBackgroundAutomatic segmentation and localization of lesions in mammogram (MG) images are challenging even with employing advanced methods such as deep learning (DL) methods. We developed a new model based on the architecture of the semantic segmentation U-Net model to precisely segment mass lesions in MG images. The proposed end-to-end convolutional neural network (CNN) based model extracts contextual information by combining low-level and high-level features. We trained the proposed model using huge publicly available databases, (CBIS-DDSM, BCDR-01, and INbreast), and a private database from the University of Connecticut Health Center (UCHC).ResultsWe compared the performance of the proposed model with those of the state-of-the-art DL models including the fully convolutional network (FCN), SegNet, Dilated-Net, original U-Net, and Faster R-CNN models and the conventional region growing (RG) method. The proposed Vanilla U-Net model outperforms the Faster R-CNN model significantly in terms of the runtime and the Intersection over Union metric (IOU). Training with digitized film-based and fully digitized MG images, the proposed Vanilla U-Net model achieves a mean test accuracy of 92.6%. The proposed model achieves a mean Dice coefficient index (DI) of 0.951 and a mean IOU of 0.909 that show how close the output segments are to the corresponding lesions in the ground truth maps. Data augmentation has been very effective in our experiments resulting in an increase in the mean DI and the mean IOU from 0.922 to 0.951 and 0.856 to 0.909, respectively.ConclusionsThe proposed Vanilla U-Net based model can be used for precise segmentation of masses in MG images. This is because the segmentation process incorporates more multi-scale spatial context, and captures more local and global context to predict a precise pixel-wise segmentation map of an input full MG image. These detected maps can help radiologists in differentiating benign and malignant lesions depend on the lesion shapes. We show that using transfer learning, introducing augmentation, and modifying the architecture of the original model results in better performance in terms of the mean accuracy, the mean DI, and the mean IOU in detecting mass lesion compared to the other DL and the conventional models.

Download Full-text

Multi-Resolution Image Segmentation Based on a Cascaded U-ADenseNet for the Liver and Tumors

Journal of Personalized Medicine ◽

10.3390/jpm11101044 ◽

2021 ◽

Vol 11 (10) ◽

pp. 1044

Author(s):

Yan Zhu ◽

Aihong Yu ◽

Huan Rong ◽

Dongqing Wang ◽

Yuqing Song ◽

...

Keyword(s):

Liver Tumors ◽

Malignant Tumors ◽

Contextual Information ◽

Automatic Segmentation ◽

Tumor Segmentation ◽

Proposed Model ◽

Wide Range ◽

Spatial Pyramid Pooling ◽

The One ◽

Coarse To Fine

The liver is an irreplaceable organ in the human body, maintaining life activities and metabolism. Malignant tumors of the liver have a high mortality rate at present. Computer-aided segmentation of the liver and tumors has significant effects on clinical diagnosis and treatment. There are still many challenges in the segmentation of the liver and liver tumors simultaneously, such as, on the one hand, that convolutional kernels with fixed geometric structures do not match complex, irregularly shaped targets; on the other, pooling during convolution results in a loss of spatial contextual information of images. In this work, we designed a cascaded U-ADenseNet with coarse-to-fine processing for addressing the above issues of fully automatic segmentation. This work contributes multi-resolution input images and multi-layered channel attention combined with atrous spatial pyramid pooling densely connected in the fine segmentation. The proposed model was evaluated by a public dataset of the Liver Tumor Segmentation Challenge (LiTS). Our approach attained competitive liver and tumor segmentation scores that exceeded other methods across a wide range of metrics.

Download Full-text

Ultrastructural Cytochemical Study of Human Gastric Cancer: Observation of AKP ase,ACP ase,G6P ase,TPP ase and CO ase

Proceedings, annual meeting, Electron Microscopy Society of America ◽

10.1017/s0424820100158819 ◽

1990 ◽

Vol 48 (3) ◽

pp. 254-255

Author(s):

Dong Yuming ◽

Yang Guanglin ◽

Du Wei Dong ◽

Xu Ai Liam

Keyword(s):

Gastric Cancer ◽

Gastric Epithelium ◽

Human Gastric Cancer ◽

Electron Microscopic ◽

Cytochemical Study ◽

Electron Microscopic Cytochemistry ◽

Cytochemical Reaction ◽

Set Up ◽

Cancer Tissues ◽

Cytochemical Technique

The activities and distributions of AKPase ,ACPase,G6Pase,TPPase and COase in human normal gastric mucosa and gastric cancer tissues were studied histochemically at light microscopic level. These enzymes are the marker enzymes of cell membrane lysosome endoplasmic reticulum, Golgi apparatus and mitochondrion objectively. On the basis of the research we set up a special ultrastructural cytochemical technique and first researched into gastric cancer domesticly. Ultrastructural cytochemistry is also called electron microscopic cytochemistry. This new technique possesses both the sensitivity of cytochemical reaction andi the high resolution of electron microscope. It is characterized by direct observation,exact localization and the combination morphology with function.The distributions of AKPase,ACPase,G6Pase,TPPase and COase in 14 cases of gastric cancer and 1 case of gastric Denign lesion were studied ultrastructurally. The results showed: 1. normal gastric epithelium had no AKPase reaction. The reaction of ACPase,G6Pase,TPPase and Coase were found in the corresponding organella, which were consistent with their function.

Download Full-text

Ultrastructural Localization Study of Carcinoembryonic Antigen (CEA) in Gastric Cancer: Immunoelectron Microscopic Observation

Proceedings, annual meeting, Electron Microscopy Society of America ◽

10.1017/s0424820100158807 ◽

1990 ◽

Vol 48 (3) ◽

pp. 252-253

Author(s):

Dong Yuming ◽

Yang Guanglin ◽

Wu Jifeng ◽

Chen Xiaolin

Keyword(s):

Gastric Cancer ◽

Intestinal Metaplasia ◽

Cancer Cells ◽

Microscopic Observation ◽

Biological Significance ◽

Ultrastructural Localization ◽

Mucinous Carcinoma ◽

Surface Glycoprotein ◽

Tubular Adenocarcinoma ◽

Pap Method

On the basis of light microscopic observation, the ultrastructural localization of CEA in gastric cancer was studied by immunoelectron microscopic technique. The distribution of CEA in gastric cancer and its biological significance and the mechanism of abnormal distribution of CEA were further discussed.Among 104 surgically resected specimens of gastric cancer with PAP method at light microscopic level, the incidence of CEA(+) was 85.58%. All of mucinous carcinoma exhibited CEA(+). In tubular adenocarcinoma the incidence of CEA(+) showed a tendency to rising with the increase of degree of differentiation. In normal epithelia and intestinal metaplasia CEA was faintly present and was found only in the luminal surface. The CEA staining patterns in cancer cells were of three types--- cytoplasmic, membranous and weak reactive type. The ultrastructural localization of CEA in 14 cases of gastric cancer was studied by immunoelectron microscopic technique.There was a little or no CEA in the microvilli of normal epithelia. In intestinal metaplasia CEA was found on the microvilli of absorptive cells and among the mucus particles of goblet cells. In gastric cancer CEA was also distributed on the lateral and basal surface or even over the entire surface of cancer cells and lost their polarity completely. Many studies had proved that the alterations in surface glycoprotein were characteristic changes of tumor cells. The antigenic determinant of CEA was glycoprotein, so the alterations of tumor-associated surface glycoprotein opened up a new way for the diagnosis of tumors.

Download Full-text