A deep learning method for automatic segmentation of the bony orbit in MRI and CT images

AbstractThis paper proposes a fully automatic method to segment the inner boundary of the bony orbit in two different image modalities: magnetic resonance imaging (MRI) and computed tomography (CT). The method, based on a deep learning architecture, uses two fully convolutional neural networks in series followed by a graph-search method to generate a boundary for the orbit. When compared to human performance for segmentation of both CT and MRI data, the proposed method achieves high Dice coefficients on both orbit and background, with scores of 0.813 and 0.975 in CT images and 0.930 and 0.995 in MRI images, showing a high degree of agreement with a manual segmentation by a human expert. Given the volumetric characteristics of these imaging modalities and the complexity and time-consuming nature of the segmentation of the orbital region in the human skull, it is often impractical to manually segment these images. Thus, the proposed method provides a valid clinical and research tool that performs similarly to the human observer.

Download Full-text

Automatic Segmentation and Classification of COVID-19 CT Image Using Deep Learning and Multi-Scale Recurrent Neural Network Based Classifier

Journal of Medical Imaging and Health Informatics ◽

10.1166/jmihi.2021.3850 ◽

2021 ◽

Vol 11 (10) ◽

pp. 2618-2625

Author(s):

R. T. Subhalakshmi ◽

S. Appavu Alias Balamurugan ◽

S. Sasikala

Keyword(s):

Neural Network ◽

Feature Extraction ◽

Deep Learning ◽

Recurrent Neural Network ◽

Automatic Segmentation ◽

Ct Images ◽

Superior Performance ◽

Multi Scale ◽

Proposed Model ◽

Class Labels

In recent times, the COVID-19 epidemic turn out to be increased in an extreme manner, by the accessibility of an inadequate amount of rapid testing kits. Consequently, it is essential to develop the automated techniques for Covid-19 detection to recognize the existence of disease from the radiological images. The most ordinary symptoms of COVID-19 are sore throat, fever, and dry cough. Symptoms are able to progress to a rigorous type of pneumonia with serious impediment. As medical imaging is not recommended currently in Canada for crucial COVID-19 diagnosis, systems of computer-aided diagnosis might aid in early COVID-19 abnormalities detection and help out to observe the disease progression, reduce mortality rates potentially. In this approach, a deep learning based design for feature extraction and classification is employed for automatic COVID-19 diagnosis from computed tomography (CT) images. The proposed model operates on three main processes based pre-processing, feature extraction, and classification. The proposed design incorporates the fusion of deep features using GoogLe Net models. Finally, Multi-scale Recurrent Neural network (RNN) based classifier is applied for identifying and classifying the test CT images into distinct class labels. The experimental validation of the proposed model takes place using open-source COVID-CT dataset, which comprises a total of 760 CT images. The experimental outcome defined the superior performance with the maximum sensitivity, specificity, and accuracy.

Download Full-text

Interobserver and intraobserver comparison of imaging glenoid morphology, glenoid version and humeral head subluxation

Shoulder & Elbow ◽

10.1177/1758573218768507 ◽

2018 ◽

Vol 11 (3) ◽

pp. 204-209 ◽

Cited By ~ 1

Author(s):

Paul J Cagle ◽

Birgit Werner ◽

Dave R Shukla ◽

Daniel A London ◽

Bradford O Parsons ◽

...

Keyword(s):

Humeral Head ◽

Interobserver Agreement ◽

Weighted Kappa ◽

Similar Efficacy ◽

Ct Images ◽

Glenoid Version ◽

Kappa Value ◽

Glenoid Morphology ◽

Magnetic Resonance Imaging Mri ◽

Ct And Mri

Background Glenoid morphology, glenoid version and humeral head subluxation represent important parameters for the treating physician. The most common method of assessing glenoid morphology is the Walch classification which has only been validated with computed tomography (CT). Methods CT images and magnetic resonance imaging (MRI) images of 25 patients were de-identified and randomized. Three reviewers assessed the images for each parameter twice. The Walch classification was assessed with a weighted kappa value. Glenoid version and humeral head subluxation were comparted with a reproducibility coefficient. Results The Walch classification demonstrated almost perfect intraobserver agreement for MRI and CT images (k = 0.87). Weighted interobserver agreement values for the Walch classification were fair for CT and MRI (k = 0.34). The weighted reproducibility coefficient for glenoid version measured 9.13 (CI 7.16–12.60) degrees for CT and 13.44 (CI 10.54–18.55) degrees for MRI images. The weighted reproducibility coefficient for percentage of humeral head subluxation was 17.43% (CI 13.67–24.06) for CT and 18.49% (CI 14.5–25.52) for MRI images. Discussion CT and MRI images demonstrated similar efficacy in classifying glenoid morphology, measuring glenoid version and measuring posterior humeral head subluxation. MRI can be used as an alternative to CT for measuring these parameters.

Download Full-text

Performance evaluation of 2D and 3D deep learning approaches for automatic segmentation of multiple organs on CT images

Medical Imaging 2018: Computer-Aided Diagnosis ◽

10.1117/12.2295178 ◽

2018 ◽

Cited By ~ 6

Author(s):

Xiangrong Zhou ◽

Kuzuma Yamada ◽

Ryosuke Takayama ◽

Xinxin Zhou ◽

Takeshi Hara ◽

...

Keyword(s):

Performance Evaluation ◽

Deep Learning ◽

Automatic Segmentation ◽

Ct Images ◽

Learning Approaches ◽

Multiple Organs ◽

2D And 3D

Download Full-text

3D U-Net for Skull Stripping in Brain MRI

Applied Sciences ◽

10.3390/app9030569 ◽

2019 ◽

Vol 9 (3) ◽

pp. 569 ◽

Cited By ~ 14

Author(s):

Hyunho Hwang ◽

Hafiz Zia Ur Rehman ◽

Sungon Lee

Keyword(s):

Deep Learning ◽

Automatic Segmentation ◽

Brain Mri ◽

Brain Magnetic Resonance Imaging ◽

Brain Extraction ◽

Skull Stripping ◽

Segmentation Algorithms ◽

Magnetic Resonance Imaging Mri ◽

Deep Learning Network ◽

The Brain

Skull stripping in brain magnetic resonance imaging (MRI) is an essential step to analyze images of the brain. Although manual segmentation has the highest accuracy, it is a time-consuming task. Therefore, various automatic segmentation algorithms of the brain in MRI have been devised and proposed previously. However, there is still no method that solves the entire brain extraction problem satisfactorily for diverse datasets in a generic and robust way. To address these shortcomings of existing methods, we propose the use of a 3D-UNet for skull stripping in brain MRI. The 3D-UNet was recently proposed and has been widely used for volumetric segmentation in medical images due to its outstanding performance. It is an extended version of the previously proposed 2D-UNet, which is based on a deep learning network, specifically, the convolutional neural network. We evaluated 3D-UNet skull-stripping using a publicly available brain MRI dataset and compared the results with three existing methods (BSE, ROBEX, and Kleesiek’s method; BSE and ROBEX are two conventional methods, and Kleesiek’s method is based on deep learning). The 3D-UNet outperforms two typical methods and shows comparable results with the specific deep learning-based algorithm, exhibiting a mean Dice coefficient of 0.9903, a sensitivity of 0.9853, and a specificity of 0.9953.

Download Full-text

Automatic Segmentation of the Prostate on 3D CT Images by Using Multiple Deep Learning Networks

Proceedings of the 2018 5th International Conference on Biomedical and Bioinformatics Engineering - ICBBE '18 ◽

10.1145/3301879.3301883 ◽

2018 ◽

Author(s):

Jiayang Xiong ◽

Luan Jiang ◽

Qiang li

Keyword(s):

Deep Learning ◽

Automatic Segmentation ◽

Ct Images ◽

3D Ct ◽

Learning Networks

Download Full-text

Hybrid 3D-ResNet Deep Learning Model for Automatic Segmentation of Thoracic Organs at Risk in CT Images

2020 International Conference on Industrial Engineering, Applications and Manufacturing (ICIEAM) ◽

10.1109/icieam48468.2020.9111950 ◽

2020 ◽

Author(s):

Abdul Qayyum ◽

Chun Kit Ang ◽

S. Sridevi ◽

M. K. A. Ahamed Khan ◽

Lim Wei Hong ◽

...

Keyword(s):

At Risk ◽

Deep Learning ◽

Organs At Risk ◽

Automatic Segmentation ◽

Learning Model ◽

Ct Images ◽

Deep Learning Model

Download Full-text

Diagnosis of COVID-19 Using a Deep Learning Model in Various Radiology Domains

Complexity ◽

10.1155/2021/1296755 ◽

2021 ◽

Vol 2021 ◽

pp. 1-10

Author(s):

Yousef Alhwaiti ◽

Muhammad Hameed Siddiqi ◽

Madallah Alruwaili ◽

Ibrahim Alrashdi ◽

Saad Alanazi ◽

...

Keyword(s):

Deep Learning ◽

Weighted Average ◽

Learning Model ◽

Front Line ◽

X Ray ◽

Proposed Model ◽

X Ray Computed ◽

Magnetic Resonance Imaging Mri ◽

Deep Learning Model ◽

Ct And Mri

Many countries are severely affected by COVID-19, and various casualties have been reported. Most countries have implemented full and partial lockdowns to control COVID-19. Paramedical employee infections are always a threatening discovery. Front-line paramedical employees might initially be at risk when observing and treating patients, who can contaminate them through respiratory secretions. If proper preventive measures are absent, front-line paramedical workers will be in danger of contamination and can become unintentional carriers to patients admitted in the hospital for other illnesses and treatments. Moreover, every country has limited testing capacity; therefore, a system is required which helps the doctor to directly check and analyze the patients’ blood structure. This study proposes a generalized adaptive deep learning model that helps the front-line paramedical employees to easily detect COVID-19 in different radiology domains. In this work, we designed a model using convolutional neural network in order to detect COVID-19 from X-ray, Computed Tomography (CT), and Magnetic Resonance Imaging (MRI) images. The proposed model has 27 layers (input, convolutional, max-pooling, dropout, flatten, dense, and output layers), which has been tested and validated on various radiology domains such as X-ray, CT, and MRI. For experiments, we utilized 70% of the dataset for training and 30% for testing against each dataset. The weighted average accuracies for the proposed model are 94%, 85%, and 86% on X-ray, CT, and MRI, respectively. The experiments show the significance of the model against state-of-the-art works.

Download Full-text

Automatic Segmentation of Pelvic Cancers Using Deep Learning: State-of-the-Art Approaches and Challenges

Diagnostics ◽

10.3390/diagnostics11111964 ◽

2021 ◽

Vol 11 (11) ◽

pp. 1964

Author(s):

Reza Kalantar ◽

Gigin Lin ◽

Jessica M. Winfield ◽

Christina Messiou ◽

Susan Lalondrelle ◽

...

Keyword(s):

Deep Learning ◽

Medical Image ◽

State Of The Art ◽

Technological Development ◽

Automatic Segmentation ◽

Research Attention ◽

Segmentation Algorithms ◽

Rectal Cancers ◽

Magnetic Resonance Imaging Mri ◽

Aided Diagnosis

The recent rise of deep learning (DL) and its promising capabilities in capturing non-explicit detail from large datasets have attracted substantial research attention in the field of medical image processing. DL provides grounds for technological development of computer-aided diagnosis and segmentation in radiology and radiation oncology. Amongst the anatomical locations where recent auto-segmentation algorithms have been employed, the pelvis remains one of the most challenging due to large intra- and inter-patient soft-tissue variabilities. This review provides a comprehensive, non-systematic and clinically-oriented overview of 74 DL-based segmentation studies, published between January 2016 and December 2020, for bladder, prostate, cervical and rectal cancers on computed tomography (CT) and magnetic resonance imaging (MRI), highlighting the key findings, challenges and limitations.

Download Full-text