scholarly journals Auto-detection of strong gravitational lenses using convolutional neural networks

2018 ◽  
Vol 2 ◽  
pp. 1 ◽  
Author(s):  
James Pearson ◽  
Clara Pennock ◽  
Tom Robinson
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Convolutional Neural Network ◽  
Area Under Curve ◽  
Operating Characteristic ◽  
Characteristic Curve ◽  
Automated Detection ◽  
Gravitational Lenses ◽  
Simulated Image ◽  
Operating Characteristic Curve

We propose a method for the automated detection of strong galaxy-galaxy gravitational lenses in images, utilising a convolutional neural network (CNN) trained on 210 000 simulated galaxy-galaxy lens and non-lens images. The CNN, named LensFinder, was tested on a separate 210 000 simulated image catalogue, with 95% of images classied with at least 98.6% certainty. An accuracy of over 98% was achieved and an area under curve of 0.9975 was determined from the resulting receiver operating characteristic curve. A regional CNN, R-LensFinder, was trained to label lens positions in images, perfectly labelling 80% while partially labelling another 10% correctly.

scholarly journals Development and Evaluation of an AI System for COVID-19 Diagnosis

2020 ◽  
Author(s):  
Cheng Jin ◽  
Weixiang Chen ◽  
Yukun Cao ◽  
Zhanwei Xu ◽  
Xin Zhang ◽  
...  
Keyword(s):  
Neural Network ◽  
Artificial Intelligence ◽  
Early Detection ◽  
Convolutional Neural Network ◽  
Operating Characteristic ◽  
Characteristic Curve ◽  
Chest Ct ◽  
Large Dataset ◽  
Operating Characteristic Curve ◽  
Timely Treatment

AbstractEarly detection of COVID-19 based on chest CT will enable timely treatment of patients and help control the spread of the disease. With rapid spreading of COVID-19 in many countries, however, CT volumes of suspicious patients are increasing at a speed much faster than the availability of human experts. Here, we propose an artificial intelligence (AI) system for fast COVID-19 diagnosis with an accuracy comparable to experienced radiologists. A large dataset was constructed by collecting 970 CT volumes of 496 patients with confirmed COVID-19 and 260 negative cases from three hospitals in Wuhan, China, and 1,125 negative cases from two publicly available chest CT datasets. Trained using only 312 cases, our diagnosis system, which is based on deep convolutional neural network, is able to achieve an accuracy of 94.98%, an area under the receiver operating characteristic curve (AUC) of 97.91%, a sensitivity of 94.06%, and a specificity of 95.47% on an independent external verification dataset of 1,255 cases. In a reader study involving five radiologists, only one radiologist is slightly more accurate than the AI system. The AI system is two orders of magnitude faster than radiologists and the code is available at https://github.com/ChenWWWeixiang/diagnosis_covid19.

scholarly journals A Coarse-to-Fine Fully Convolutional Neural Network for Fundus Vessel Segmentation

Symmetry ◽  
2018 ◽  
Vol 10 (11) ◽  
pp. 607 ◽  
Author(s):  
Jianwei Lu ◽  
Yixuan Xu ◽  
Mingle Chen ◽  
Ye Luo
Keyword(s):  
Neural Network ◽  
Convolutional Neural Network ◽  
Operating Characteristic ◽  
Characteristic Curve ◽  
Original Data ◽  
Vessel Segmentation ◽  
Retinal Diseases ◽  
Fundus Images ◽  
The Neural Network ◽  
Coarse To Fine

Fundus vessel analysis is a significant tool for evaluating the development of retinal diseases such as diabetic retinopathy and hypertension in clinical practice. Hence, automatic fundus vessel segmentation is essential and valuable for medical diagnosis in ophthalmopathy and will allow identification and extraction of relevant symmetric and asymmetric patterns. Further, due to the uniqueness of fundus vessel, it can be applied in the field of biometric identification. In this paper, we remold fundus vessel segmentation as a task of pixel-wise classification task, and propose a novel coarse-to-fine fully convolutional neural network (CF-FCN) to extract vessels from fundus images. Our CF-FCN is aimed at making full use of the original data information and making up for the coarse output of the neural network by harnessing the space relationship between pixels in fundus images. Accompanying with necessary pre-processing and post-processing operations, the efficacy and efficiency of our CF-FCN is corroborated through our experiments on DRIVE, STARE, HRF and CHASE DB1 datasets. It achieves sensitivity of 0.7941, specificity of 0.9870, accuracy of 0.9634 and Area Under Receiver Operating Characteristic Curve (AUC) of 0.9787 on DRIVE datasets, which surpasses the state-of-the-art approaches.

scholarly journals Sensitivity and Specificity of Spectral Oct in Patients with Early Glaucoma

2019 ◽  
Vol 75 (5) ◽  
pp. 260-264
Author(s):  
Magdaléna Bočková ◽  
Petr Veselý ◽  
Svatopluk Synek ◽  
Lubomír Hanák ◽  
Pavel Beneš
Keyword(s):  
Retrospective Study ◽  
Visual Field ◽  
Sensitivity And Specificity ◽  
Area Under Curve ◽  
Operating Characteristic ◽  
Characteristic Curve ◽  
Test Statistics ◽  
Early Glaucoma ◽  
Spectral Oct ◽  
Operating Characteristic Curve

The study examines the sensitivity and specificity of spectral OCT in detecting early glaucoma. The aim was to evaluate data obtained by RNFL analysis in 4 observed quadrants and to compare it with the resulting diagnosis of glaucoma neuropathy determined subsequently on the basis of changes in the visual field. This concerns a retrospective study numbering 31 probands who underwent OCT examination at our centre in the period from 2008 to 2017. Test statistics demonstrated sensitivity of OCT examination (specific RNFL analysis) of 63.64% and specificity of 90%. The used ROC (receiver operating characteristic curve) test showed an AUC (area under curve) value of 0.734 on a statistically significant level of p = 0.0097. We therefore found that the instrument Spectral OCT SLO, with the aid of RNFL analysis, was effective in determining probands in whom development of glaucoma pathology was subsequently confirmed.

Survivability prediction of colon cancer patients using neural networks

2017 ◽  
Vol 25 (3) ◽  
pp. 878-891 ◽  
Author(s):  
Reda Al-Bahrani ◽  
Ankit Agrawal ◽  
Alok Choudhary
Keyword(s):  
Neural Networks ◽  
Colon Cancer ◽  
Deep Neural Networks ◽  
Operating Characteristic ◽  
Prediction Models ◽  
Patient Survival ◽  
Characteristic Curve ◽  
Conditional Survival ◽  
Operating Characteristic Curve ◽  
Multiple Neural Network

We utilize deep neural networks to develop prediction models for patient survival and conditional survival of colon cancer. Our models are trained and validated on data obtained from the Surveillance, Epidemiology, and End Results Program. We provide an online outcome calculator for 1, 2, and 5 years survival periods. We experimented with multiple neural network structures and found that a network with five hidden layers produces the best results for these data. Moreover, the online outcome calculator provides conditional survival of 1, 2, and 5 years after surviving the mentioned survival periods. In this article, we report an approximate 0.87 area under the receiver operating characteristic curve measurements, higher than the 0.85 reported by Stojadinovic et al.

scholarly journals Gastrin-17 Combined With CEA, CA12-5 and CA19-9 Improves the Sensitivity For the Diagnosis of Gastric Cancer

2021 ◽  
Author(s):  
Zhen Wang ◽  
Tang-ming Mo ◽  
Lei Tian ◽  
Jun-qiang Chen
Keyword(s):  
Gastric Cancer ◽  
Area Under Curve ◽  
Operating Characteristic ◽  
Characteristic Curve ◽  
Diagnostic Value ◽  
Serum Markers ◽  
Control Group ◽  
Positive Rate ◽  
Healthy Control ◽  
Operating Characteristic Curve

Abstract Background: Previous studies reported the utility of serum tumor markers (such as CEA, CA12-5 and CA19-9) and gastrin-17 in the diagnosis of gastric cancer (GC). However, the value of these serum markers for diagnosing GC is still under debate.Methods: The level of CEA, CA12-5, CA19-9 and gastrin-17 was tested in 230 GC patients and 99 healthy people. The value of the four markers for diagnosing GC was analyzed.Results: The positive rate of Gastrin-17, CEA, CA199 and CA125 were much higher in GC group (22.61%, 22.61%, 20.00% and 8.26%, respectively) than that of healthy control group (5.05%, 2.02%, 1.01% and 2.02%, respectively). The sensitivity of Gastrin-17, CEA, CA125 and CA199 in the diagnosis of GC were 22.61%, 22.61%, 6.96% and 20.00%, respectively, and the corresponding specificity were 94.95%, 97.98%, 98.99% and 98.99%, respectively. By using the optimal cut-off value derived from the area under curve (AUC) of receiver operating characteristic curve, the AUC of gastin-17, CEA, CA125, CA199 increased to 0.72, 0.64, 0.61 and 0.65, respectively. After combining the four markers, the AUC increased to 0.79 (95% CI: 0.75-0.84), and the corresponding sensitivity and specificity were 65.22% (95% CI: 58.70% - 71.40%) and 84.85% (95% CI: 76.20% - 91.30%), respectively.Conclusion: CEA, CA12-5, CA19-9 and gastrin-17 were all valuable in the diagnosis of GC, and gastrin-17 had the best diagnostic value among the four markers. Gastrin-17 combined with CEA, CA12-5 and CA19-9 could improve the diagnostic value of GC significantly. Prospective, multi-center studies are needed to validate our findings.

scholarly journals Serum miR-214 Serves as a Biomarker for Prodromal Parkinson’s Disease

2021 ◽  
Vol 13 ◽  
Author(s):  
Lanting Li ◽  
Jingru Ren ◽  
Chenxi Pan ◽  
Yuqian Li ◽  
Jianxia Xu ◽  
...  
Keyword(s):  
Parkinson’S Disease ◽  
Parkinson's Disease ◽  
Area Under Curve ◽  
Operating Characteristic ◽  
De Novo ◽  
Characteristic Curve ◽  
The Other ◽  
Circulating Micrornas ◽  
Potential Biomarker ◽  
Operating Characteristic Curve

Circulating microRNAs (miRNAs) have been proposed to be accessible biomarkers for Parkinson’s disease (PD). However, there is a lack of known miRNAs that can serve as biomarkers for prodromal PD (pPD). We previously identified that miR-31 and miR-214 were dysregulated in PD. The aim of this study was to explore the roles of miR-31 and miR-214 in pPD. We recruited 25 pPD patients, 20 patients with de novo PD (dnPD), 24 advanced PD (aPD) patients and 21 controls. Next, we investigated the expression of miR-31 and miR-214. Compared to controls, miR-214 was found to be significantly upregulated in pPD patients while miR-31 was significantly upregulated in aPD patients. In addition, the expression of miR-214 was lower in aPD patients compared to both dnPD or pPD patients, while the expression of miR-31 was higher in aPD patients compared to dnPD patients. In order to predict pPD via miRNA expression, the receiver operating characteristic curve was constructed and the area under curve (AUC) was calculated. For pPD prediction by miR-214, the AUC was 0.756. The optimal cut-off value of miR-214 was 0.1962, and the sensitivity and specificity were 72.0% and 76.2%, respectively. On the other hand, the AUC for aPD detection by miR-31 was 0.744. The optimal cut-off value for miR-31 was 0.0148, with a sensitivity of 87.5% and a specificity of 71.4%. In conclusion, miR-214 can distinguish pPD patients from controls and may be used as a potential biomarker for pPD diagnosis.

scholarly journals An AI-Based Algorithm for the Automatic Classification of Thoracic Radiographs in Cats

2021 ◽  
Vol 8 ◽  
Author(s):  
Tommaso Banzato ◽  
Marek Wodzinski ◽  
Federico Tauceri ◽  
Chiara Donà ◽  
Filippo Scavazza ◽  
...  
Keyword(s):  
Artificial Intelligence ◽  
Neural Networks ◽  
Pleural Effusion ◽  
Operating Characteristic ◽  
Characteristic Curve ◽  
Computer Aided Detection ◽  
Radiographic Findings ◽  
Computer Aided ◽  
Operating Characteristic Curve

An artificial intelligence (AI)-based computer-aided detection (CAD) algorithm to detect some of the most common radiographic findings in the feline thorax was developed and tested. The database used for training comprised radiographs acquired at two different institutions. Only correctly exposed and positioned radiographs were included in the database used for training. The presence of several radiographic findings was recorded. Consequenly, the radiographic findings included for training were: no findings, bronchial pattern, pleural effusion, mass, alveolar pattern, pneumothorax, cardiomegaly. Multi-label convolutional neural networks (CNNs) were used to develop the CAD algorithm, and the performance of two different CNN architectures, ResNet 50 and Inception V3, was compared. Both architectures had an area under the receiver operating characteristic curve (AUC) above 0.9 for alveolar pattern, bronchial pattern and pleural effusion, an AUC above 0.8 for no findings and pneumothorax, and an AUC above 0.7 for cardiomegaly. The AUC for mass was low (above 0.5) for both architectures. No significant differences were evident in the diagnostic accuracy of either architecture.

scholarly journals Pertinence of score for neonatal acute physiology-II to prognosticate mortality and organ dysfunction in neonatal sepsis

2020 ◽  
Vol 71 ◽  
pp. 60-65
Author(s):  
A. A. Verma ◽  
U. C. Rajput ◽  
A. A. Kinikar
Keyword(s):  
Organ Dysfunction ◽  
Predictive Value ◽  
Area Under Curve ◽  
Operating Characteristic ◽  
Characteristic Curve ◽  
Mortality And Morbidity ◽  
Mortality Outcome ◽  
Operating Characteristic Curve ◽  
Sensitivity Specificity ◽  
Prospective Investigation

Introduction: The present investigation was undertaken to correlation between mortality and morbidity (organ dysfunction [OD]) and score for neonatal acute physiology-II (SNAP-II). Materials and Methods: A prospective investigation of newborns neonates, a total 157 neonates 82 male (52.2%), female 75 (47.8%) were enrolled and disunited into four groups according to gestational age: 28 to 30 weeks (G1), 31 to 33 (G2) 34 to 36 weeks (G3) and >37 weeks (G4) variables analyzed were SNAP II. Results and Discussion: The receiver operating characteristic curve for SNAP-II score and death is more predictive in correlation to OD (area under curve of death is 0.776 as compared to 0.553 for OD). The sensitivity, specificity, positive predictive value, and negative predictive value of SNAP-II score with mortality (outcome) were 42.8%, 100%, 100%, and 82.3%, respectively. Conclusion: The SNAP-II revealed efficient to fantabulous ≥40 can prognosticate OD and death when applied on admission to neonates with sepsis.

Sign in / Sign up

Export Citation Format

Share Document