scholarly journals A Few-Shot U-Net Deep Learning Model for COVID-19 Infected Area Segmentation in CT Images

Sensors ◽  
2021 ◽  
Vol 21 (6) ◽  
pp. 2215
Author(s):  
Athanasios Voulodimos ◽  
Eftychios Protopapadakis ◽  
Iason Katsamenis ◽  
Anastasios Doulamis ◽  
Nikolaos Doulamis
Keyword(s):  
Deep Learning ◽  
Statistical Significance ◽  
High Sensitivity ◽  
Semantic Segmentation ◽  
Ct Images ◽  
General Concept ◽  
Fine Tuning ◽  
P Value ◽  
Training Procedure ◽  
Infected Area

Recent studies indicate that detecting radiographic patterns on CT chest scans can yield high sensitivity and specificity for COVID-19 identification. In this paper, we scrutinize the effectiveness of deep learning models for semantic segmentation of pneumonia-infected area segmentation in CT images for the detection of COVID-19. Traditional methods for CT scan segmentation exploit a supervised learning paradigm, so they (a) require large volumes of data for their training, and (b) assume fixed (static) network weights once the training procedure has been completed. Recently, to overcome these difficulties, few-shot learning (FSL) has been introduced as a general concept of network model training using a very small amount of samples. In this paper, we explore the efficacy of few-shot learning in U-Net architectures, allowing for a dynamic fine-tuning of the network weights as new few samples are being fed into the U-Net. Experimental results indicate improvement in the segmentation accuracy of identifying COVID-19 infected regions. In particular, using 4-fold cross-validation results of the different classifiers, we observed an improvement of 5.388 ± 3.046% for all test data regarding the IoU metric and a similar increment of 5.394 ± 3.015% for the F1 score. Moreover, the statistical significance of the improvement obtained using our proposed few-shot U-Net architecture compared with the traditional U-Net model was confirmed by applying the Kruskal-Wallis test (p-value = 0.026).

scholarly journals Deep learning models for COVID-19 infected area segmentation in CT images

2020 ◽  
Author(s):  
Athanasios Voulodimos ◽  
Eftychios Protopapadakis ◽  
Iason Katsamenis ◽  
Anastasios Doulamis ◽  
Nikolaos Doulamis
Keyword(s):  
Neural Networks ◽  
Deep Learning ◽  
Convolutional Neural Networks ◽  
Class Imbalance ◽  
Semantic Segmentation ◽  
Ct Images ◽  
Learning Models ◽  
Fully Convolutional Neural Networks ◽  
Infected Area ◽  
Ct Data

Recent studies indicated that detecting radiographic patterns on CT chest scans could in some cases yield higher sensitivity and specificity for COVID-19 detection compared to other methods such as RTPCR. In this work, we scrutinize the effectiveness of deep learning models for semantic segmentation of pneumonia infected area segmentation in CT images for the detection of COVID-19. We explore the efficacy of U-Nets and Fully Convolutional Neural Networks in this task using real-world CT data from COVID-19 patients. The results indicate that Fully Convolutional Neural Networks are capable of accurate segmentation despite the class imbalance on the dataset and the man-made annotation errors on the boundaries of symptom manifestation areas, and can be a promising method for further analysis of COVID-19 induced pneumonia symptoms in CT images.

scholarly journals Vulnerability in Deep Transfer Learning Models to Adversarial Fast Gradient Sign Attack for COVID-19 Prediction from Chest Radiography Images

2021 ◽  
Vol 11 (9) ◽  
pp. 4233
Author(s):  
Biprodip Pal ◽  
Debashis Gupta ◽  
Md. Rashed-Al-Mahfuz ◽  
Salem A. Alyami ◽  
Mohammad Ali Moni
Keyword(s):  
Deep Learning ◽  
Transfer Learning ◽  
Chest Radiography ◽  
High Sensitivity ◽  
Ct Images ◽  
Classification Performance ◽  
Learning Models ◽  
X Ray ◽  
Image Capturing ◽  
Fast Gradient

The COVID-19 pandemic requires the rapid isolation of infected patients. Thus, high-sensitivity radiology images could be a key technique to diagnose patients besides the polymerase chain reaction approach. Deep learning algorithms are proposed in several studies to detect COVID-19 symptoms due to the success in chest radiography image classification, cost efficiency, lack of expert radiologists, and the need for faster processing in the pandemic area. Most of the promising algorithms proposed in different studies are based on pre-trained deep learning models. Such open-source models and lack of variation in the radiology image-capturing environment make the diagnosis system vulnerable to adversarial attacks such as fast gradient sign method (FGSM) attack. This study therefore explored the potential vulnerability of pre-trained convolutional neural network algorithms to the FGSM attack in terms of two frequently used models, VGG16 and Inception-v3. Firstly, we developed two transfer learning models for X-ray and CT image-based COVID-19 classification and analyzed the performance extensively in terms of accuracy, precision, recall, and AUC. Secondly, our study illustrates that misclassification can occur with a very minor perturbation magnitude, such as 0.009 and 0.003 for the FGSM attack in these models for X-ray and CT images, respectively, without any effect on the visual perceptibility of the perturbation. In addition, we demonstrated that successful FGSM attack can decrease the classification performance to 16.67% and 55.56% for X-ray images, as well as 36% and 40% in the case of CT images for VGG16 and Inception-v3, respectively, without any human-recognizable perturbation effects in the adversarial images. Finally, we analyzed that correct class probability of any test image which is supposed to be 1, can drop for both considered models and with increased perturbation; it can drop to 0.24 and 0.17 for the VGG16 model in cases of X-ray and CT images, respectively. Thus, despite the need for data sharing and automated diagnosis, practical deployment of such program requires more robustness.

scholarly journals A Medical Service Application Based on 3D-CNN and Knowledge Graph

2021 ◽  
Vol 2078 (1) ◽  
pp. 012048
Author(s):  
Jiasheng Ni
Keyword(s):  
Lung Cancer ◽  
Deep Learning ◽  
Cancer Detection ◽  
Semantic Segmentation ◽  
Ct Images ◽  
Knowledge Graph ◽  
Medical Response ◽  
Learning Techniques ◽  
3D Cnn ◽  
Lung Cancer Detection

Abstract Remote medical prognosis application is a category of medical tests tool designed to collect patients’ body conditions and offer diagnosis results synchronously. However, most online applications are predicated on a simple chat bot which typically redirect patients to other online medical websites, which undermines the user experience and may prompt useless information for their reference. To tackle these issues, this paper proposed a medical prognosis application with deep learning techniques for a more responsive and intelligent medical prognosis procedure. This application can be break down into three parts-lung cancer detection, a database-supporting medical QA bot and a Hierarchical Bidirectional LSTM model (HBDA). A 3D-CNN model is built for the lung cancer detection, with a sequence of sliced CT images as inputs and outputs a probability scaler for tumor indications. A knowledge graph is applied in the medical QA bot implementation and the HBDA model is designed for semantic segmentation in order to better capture users’ intention in medical questions. For the performance of the medical prognosis, since we have limited computer memory, the 3D-CNN didn’t perform very well on detecting tumor conditions in the CT images with accuracy at around 70%. The knowledge graph-based medical QA bot intelligently recognize the underlying pattern in patients’ question and delivered decent medical response. The HBDA model performs well on distinguish the similarities and disparities between various medical questions, reaching accuracy at 90%. These results shed light for the feasibility of utilizing deep learning techniques such as 3D-CNN, Knowledge Graph, and Hierarchical Bi-directional LSTM to simulate the medical prognosis process.

scholarly journals Deep learning for plankton and coral classification

2020 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Author(s):  
Alessandra Lumini ◽  
Loris Nanni ◽  
Gianluca Maguolo
Keyword(s):  
Deep Learning ◽  
Performance Improvement ◽  
State Of The Art ◽  
Source Code ◽  
Fine Tuning ◽  
Automated System ◽  
Training Procedure ◽  
Single Model ◽  
Training Models ◽  
Heterogeneous Ensemble

In this paper, we present a study about an automated system for monitoring underwater ecosystems. The system here proposed is based on the fusion of different deep learning methods. We study how to create an ensemble based of different Convolutional Neural Network (CNN) models, fine-tuned on several datasets with the aim of exploiting their diversity. The aim of our study is to experiment the possibility of fine-tuning CNNs for underwater imagery analysis, the opportunity of using different datasets for pre-training models, the possibility to design an ensemble using the same architecture with small variations in the training procedure.Our experiments, performed on 5 well-known datasets (3 plankton and 2 coral datasets) show that the combination of such different CNN models in a heterogeneous ensemble grants a substantial performance improvement with respect to other state-of-the-art approaches in all the tested problems. One of the main contributions of this work is a wide experimental evaluation of famous CNN architectures to report the performance of both the single CNN and the ensemble of CNNs in different problems. Moreover, we show how to create an ensemble which improves the performance of the best single model. The MATLAB source code is freely link provided in title page.

An international challenge to use artificial intelligence to define the state-of-the-art in kidney and kidney tumor segmentation in CT imaging.

2020 ◽  
Vol 38 (6_suppl) ◽  
pp. 626-626
Author(s):  
Nicholas Heller ◽  
Sean McSweeney ◽  
Matthew Thomas Peterson ◽  
Sarah Peterson ◽  
Jack Rickman ◽  
...  
Keyword(s):  
Deep Learning ◽  
Treatment Decision ◽  
Semantic Segmentation ◽  
Ct Images ◽  
Renal Tumors ◽  
Tumor Segmentation ◽  
Computer Assisted ◽  
Kidney Tumor ◽  
3D Segmentation ◽  
Kidney Tumors

626 Background: The 2019 Kidney and Kidney Tumor Segmentation challenge (KiTS19) was an international competition held in conjunction with the 2019 International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI) and sought to stimulate progress on this automatic segmentation frontier. Growing rates of kidney tumor incidence led to research into the use of artificial inteligence (AI) to radiographically differentiate and objectively characterize these tumors. Automated segmentation using AI objectively quantifies complexity and aggression of renal tumors to better differentiate and describe the tumors for improved treatment decision making. Methods: A training set of over 31,000 CT images from 210 patients with kidney tumors was publicly released with corresponding semantic segmentation masks. 106 teams from five continents used this data to develop automated deep learning systems to predict the true segmentation masks on a test set of an additional 13,500 CT images in 90 patients for which the corresponding ground truth segmentations were kept private. These predictions were scored and ranked according to their average Sørensen-Dice coefficient between kidney and tumor across the 90 test cases. Results: The winning team achieved a Dice of 0.974 for kidney and 0.851 for tumor, approaching the human inter-annotator performance on kidney (0.983) but falling short on tumor (0.923). This challenge has now entered an “open leaderboard” phase where it serves as a challenging benchmark in 3D semantic segmentation. Conclusions: Results of the KiTS19 challenge show deep learning methods are fully capable of reliable segmentation of kidneys and kidney tumors. The KiTS19 challenge attracted a high number of submissions and serves as an important and challenging benchmark in 3D segmentation. The publicly available data will further propel the use of automated 3D segmentation analysis. Fully segmented kidneys and tumors allow for automated calculation of all types of nephrometry, tumor textural variation and discovery of new predictive features important for personalized medicine and accurate prediction of patient relevant outcomes.

Evaluation of a New Neural Network Classifier for Diabetic Retinopathy

2021 ◽  
pp. 193229682110426
Author(s):  
Or Katz ◽  
Dan Presil ◽  
Liz Cohen ◽  
Roi Nachmani ◽  
Naomi Kirshner ◽  
...  
Keyword(s):  
Diabetic Retinopathy ◽  
Deep Learning ◽  
Screening Program ◽  
Semantic Segmentation ◽  
Medical Image Segmentation ◽  
Grading System ◽  
Sufficient Information ◽  
Training Procedure ◽  
Screening Process ◽  
Diabetic Eye Screening

Background: Medical image segmentation is a well-studied subject within the field of image processing. The goal of this research is to create an AI retinal screening grading system that is both accurate and fast. We introduce a new segmentation network which achieves state-of-the-art results on semantic segmentation of color fundus photographs. By applying the net-work to identify anatomical markers of diabetic retinopathy (DR) and diabetic macular edema (DME), we collect sufficient information to classify patients by grades R0 and R1 or above, M0 and M1. Methods: The AI grading system was trained on screening data to evaluate the presence of DR and DME. The core algorithm of the system is a deep learning network that segments relevant anatomical features in a retinal image. Patients were graded according to the standard NHS Diabetic Eye Screening Program feature-based grading protocol. Results: The algorithm performance was evaluated with a series of 6,981 patient retinal images from routine diabetic eye screenings. It correctly predicted 98.9% of retinopathy events and 95.5% of maculopathy events. Non-disease events prediction rate was 68.6% for retinopathy and 81.2% for maculopathy. Conclusion: This novel deep learning model was trained and tested on patient data from annual diabetic retinopathy screenings can classify with high accuracy the DR and DME status of a person with diabetes. The system can be easily reconfigured according to any grading protocol, without running a long AI training procedure. The incorporation of the AI grading system can increase the graders’ productivity and improve the final outcome accuracy of the screening process.

scholarly journals Deep-Learning-Based Gridded Downscaling of Surface Meteorological Variables in Complex Terrain. Part I: Daily Maximum and Minimum 2-m Temperature

2020 ◽  
Vol 59 (12) ◽  
pp. 2057-2073
Author(s):  
Yingkai Sha ◽  
David John Gagne II ◽  
Gregory West ◽  
Roland Stull
Keyword(s):  
United States ◽  
Deep Learning ◽  
Expert Knowledge ◽  
Semantic Segmentation ◽  
Absolute Error ◽  
Fine Tuning ◽  
Daily Maximum ◽  
Fine Grained ◽  
Level Performance ◽  
Maximum Minimum

AbstractMany statistical downscaling methods require observational inputs and expert knowledge and thus cannot be generalized well across different regions. Convolutional neural networks (CNNs) are deep-learning models that have generalization abilities for various applications. In this research, we modify UNet, a semantic-segmentation CNN, and apply it to the downscaling of daily maximum/minimum 2-m temperature (TMAX/TMIN) over the western continental United States from 0.25° to 4-km grid spacings. We select high-resolution (HR) elevation, low-resolution (LR) elevation, and LR TMAX/TMIN as inputs; train UNet using Parameter–Elevation Regressions on Independent Slopes Model (PRISM) data over the south- and central-western United States from 2015 to 2018; and test it independently over both the training domains and the northwestern United States from 2018 to 2019. We found that the original UNet cannot generate enough fine-grained spatial details when transferred to the new northwestern U.S. domain. In response, we modified the original UNet by assigning an extra HR elevation output branch/loss function and training the modified UNet to reproduce both the supervised HR TMAX/TMIN and the unsupervised HR elevation. This improvement is named “UNet-Autoencoder (AE).” UNet-AE supports semisupervised model fine-tuning for unseen domains and showed better gridpoint-level performance with more than 10% mean absolute error (MAE) reduction relative to the original UNet. On the basis of its performance relative to the 4-km PRISM, UNet-AE is a good option to provide generalizable downscaling for regions that are underrepresented by observations.

scholarly journals LESION DETECTION IN CT IMAGES USING DEEP LEARNING SEMANTIC SEGMENTATION TECHNIQUE

2017 ◽  
Vol XLII-2/W4 ◽  
pp. 13-17 ◽  
Author(s):  
A. Kalinovsky ◽  
V. Liauchuk ◽  
A. Tarasau
Keyword(s):  
Deep Learning ◽  
Semantic Segmentation ◽  
Ct Images ◽  
Lesion Detection ◽  
Superior Performance ◽  
Ct Scans ◽  
Modern Concept ◽  
3D Ct ◽  
Convolutional Networks ◽  
2D Images

In this paper, the problem of automatic detection of tuberculosis lesion on 3D lung CT images is considered as a benchmark for testing out algorithms based on a modern concept of Deep Learning. For training and testing of the algorithms a domestic dataset of 338 3D CT scans of tuberculosis patients with manually labelled lesions was used. The algorithms which are based on using Deep Convolutional Networks were implemented and applied in three different ways including slice-wise lesion detection in 2D images using semantic segmentation, slice-wise lesion detection in 2D images using sliding window technique as well as straightforward detection of lesions via semantic segmentation in whole 3D CT scans. The algorithms demonstrate superior performance compared to algorithms based on conventional image analysis methods.

scholarly journals Deep learning models for COVID-19 infected area segmentation in CT images

2021 ◽  
Author(s):  
Athanasios Voulodimos ◽  
Eftychios Protopapadakis ◽  
Iason Katsamenis ◽  
Anastasios Doulamis ◽  
Nikolaos Doulamis
Keyword(s):  
Deep Learning ◽  
Ct Images ◽  
Learning Models ◽  
Infected Area
Sign in / Sign up

Export Citation Format

Share Document