scholarly journals Comparative Study of Deep Learning Models for Automatic Coronary Stenosis Detection in X-ray Angiography

2020 ◽  
pp. paper75-1-paper75-11
Author(s):  
Viacheslav Danilov ◽  
Olga Gerget ◽  
Kirill Klyshnikov ◽  
Evgeny Ovcharenko ◽  
Alejandro Frangi
Keyword(s):  
Deep Learning ◽  
Coronary Artery ◽  
Coronary Angiography ◽  
Model Complexity ◽  
Validation Dataset ◽  
Learning Models ◽  
Training Time ◽  
X Ray ◽  
Machine Learning Approach ◽  
Artery Disease

The article explores the application of machine learning approach to detect both single-vessel and multivessel coronary artery disease from X-ray angiography. Since the interpretation of coronary angiography images requires interventional cardiologists to have considerable training, our study is aimed at analysing, training, and assessing the potential of the existing object detectors for classifying and detecting coronary artery stenosis using angiographic imaging series. 100 patients who underwent coronary angiography at the Research Institute for Complex Issues of Cardiovascular Diseases were retrospectively enrolled in the study. To automate the medical data analysis, we examined and compared three models (SSD MobileNet V1, Faster-RCNN ResNet-50 V1, FasterRCNN NASNet) with various architecture, network complexity, and a number of weights. To compare developed deep learning models, we used the mean Average Precision (mAP) metric, training time, and inference time. Testing results show that the training/inference time is directly proportional to the model complexity. Thus, Faster-RCNN NASNet demonstrates the slowest inference time. Its mean inference time per one image made up 880 ms. In terms of accuracy, FasterRCNN ResNet-50 V1 demonstrates the highest prediction accuracy. This model has reached the mAP metric of 0.92 on the validation dataset. SSD MobileNet V1 has demonstrated the best inference time with the inference rate of 23 frames per second.

Abstract 2805: Assessment of Coronary Artery Disease Using Magnetic Resonance Coronary Angiography: A Multicenter Trial

Circulation ◽  
2008 ◽  
Vol 118 (suppl_18) ◽  
Author(s):  
Shingo Kato ◽  
Hajime Sakuma ◽  
Nanaka Ishida ◽  
Masaki Ishida ◽  
Motonori Nagata ◽  
...  
Keyword(s):  
Coronary Artery Disease ◽  
Coronary Artery ◽  
Magnetic Resonance ◽  
Coronary Angiography ◽  
Multicenter Study ◽  
Sublingual Administration ◽  
X Ray ◽  
Coronary Mra ◽  
Artery Disease ◽  
Roc Area

Background: CT coronary angiography is widely used to assess the presence of significant coronary artery disease (CAD). However, CT approach is associated with low but nonnegligible cancer risk. The purpose of this prospective multicenter study was to evaluate the diagnostic performance of coronary magnetic resonance angiography (MRA) in the ability to identify patients with significant CAD compared with coronary angiography. Materials and Methods: The subjects were recruited from 7 institutions. Free breathing coronary MR angiograms covering the entire coronary artery tree were obtained in 138 patients who were suspicious of CAD. Non-contrast enhanced images were acquired with a commercial 1.5T MR imager and five-element cardiac coils after sublingual administration of isosorbide dinitrate. Conventional X-ray coronary angiography was performed within 4 weeks after coronary MRA. MR and X-ray angiograms were sent to a core laboratory for blinded interpretation. Coronary MR angiograms were evaluated by two experienced investigators by using sliding partial MIP reconstruction. Quantitative X-ray coronary angiography analysis was performed with significant CAD defined as luminal narrowing of at least 50% of the diameter. Results: The mean imaging time of coronary MRA was 9.5 ± 4.9 minutes. The prevalence of significant disease on X-ray angiography was 45% (62/138). On a vessel-based analysis, the area under receiver operating characteristic (ROC) curve for the MRA compared with X-ray angiography was 0.90 (95% CI; 0.86 to 0.93). On a patient based analysis, the ROC area was 0.88 (95% CI; 0.81– 0.93). The sensitivity, specificity, positive and negative predictive values of coronary MRA by vessel analysis were 78% (95% CI; 68 – 86%), 86% (82–90%), 60% (51– 69%), 94% (90–96%). These values by patient analysis were 87% (95% CI; 76–94%), 71% (59 – 81%), 71% (59 – 81%), 87% (76–94%). Conclusions: In the current multicenter study using commercial 1.5T MR imagers and sliding partial MIP reconstruction, the diagnostic accuracy of coronary MRA compared to quantitative coronary angiography is good, reflected by an ROC area of 0.88 on patient-based analysis. High negative predictive value indicates that coronary MRA can be used for screening CAD.

scholarly journals Coronary Artery Dominance in Western Maharashtra Population by Angiographic Method

2020 ◽  
Vol 11 (SPL4) ◽  
pp. 1998-2002
Author(s):  
Sheela D. Kadam ◽  
Abhijeet Shelke ◽  
Priya P Roy ◽  
Megha A Doshi ◽  
Shruti P Mohite
Keyword(s):  
Coronary Artery Disease ◽  
Coronary Artery ◽  
Coronary Angiography ◽  
X Ray ◽  
Invasive Angiography ◽  
The World ◽  
History Of ◽  
Artery Disease ◽  
Angiographic Method ◽  
Day By Day

Coronary artery disease (CAD) is going to become a significant cause of death in the world. The CAD is increasing day by day because of the changing lifestyle of people. The responsible factors for CAD are diabetes, hypertension, addiction and heredity also. So, the present work is undertaken to study the dominant pattern of coronary artery in the Maharashtra population. The present study was a hospital-based, prospective and observational study of 360 patients who have coronary artery disease undergoing coronary angiography.  This study carried out from May 2018-November 2019 of both genders of 25 years of age and above [Male-215(59.72%) and female was 145(40.27 %)]. Out of that, the youngest patient below 40 years male was 30(73.17%), and the female was 11(26. 82%). While above 40 years males were 185(57.99%) and females was 134(42%). The patients with a history of by-pass surgery and angioplasty were excluded. Invasive angiography was performed by either femoral or radial artery using radio-opaque dye (OMNIPCK-50ml) that is visible by an x-ray machine (GE. INNOVA). The socio-demographic Proforma of patients, the pattern of coronary arterial dominance were recorded. Right coronary artery was dominant in 273(75.83%) patients observed in this study. While LCx was dominant in 49(13.6%) and co-dominant in 38 (10.55%) patients. Knowledge of study can be helpful to cardiologists for anatomical assessment of coronary arteries for diagnostic purposes and invasive studies.

scholarly journals Main Coronary Vessel Segmentation Using Deep Learning in Smart Medical

2020 ◽  
Vol 2020 ◽  
pp. 1-9
Author(s):  
Zhanchao Xian ◽  
Xiaoqing Wang ◽  
Shaodi Yan ◽  
Dahao Yang ◽  
Junyu Chen ◽  
...  
Keyword(s):  
Deep Learning ◽  
Coronary Artery ◽  
Automatic Segmentation ◽  
Disease Diagnosis ◽  
Vessel Segmentation ◽  
X Ray ◽  
Convolutional Networks ◽  
Fully Convolutional Networks ◽  
Main Vessel ◽  
Artery Disease

The automatic segmentation of main vessels on X-ray angiography (XRA) images is of great importance in the smart coronary artery disease diagnosis system. However, existing methods have been developed to this task, but these methods have difficulty in recognizing the coronary artery structure in XRA images. Main vessel segmentation is still a challenging task due to the diversity and small-size region of the vessel in the XRA images. In this study, we propose a robust method for main vessel segmentation by using deep learning architectures with fully convolutional networks. Four deep learning models based on the UNet architecture are evaluated on a clinical dataset, which consists of 3200 X-ray angiography images collected from 1118 patients. Using the precision (Pre), recall (Re), and F1 score (F1) as evaluation metrics, the average Pre, Re, and F1 for main vessel segmentation in the entire experimental dataset is 0.901, 0.898, and 0.900, respectively. 89.8% of the images exhibited a high F1 score >0.8. For the main vessel segmentation in XRA images, our deep learning methods demonstrated that vessels could be segmented in real time with a more optimized implementation, to further facilitate the online diagnosis in smart medical.

scholarly journals Transfer Learning for Stenosis Detection in X-ray Coronary Angiography

Mathematics ◽  
2020 ◽  
Vol 8 (9) ◽  
pp. 1510
Author(s):  
Emmanuel Ovalle-Magallanes ◽  
Juan Gabriel Avina-Cervantes ◽  
Ivan Cruz-Aceves ◽  
Jose Ruiz-Pinales
Keyword(s):  
Coronary Artery ◽  
Coronary Angiography ◽  
Transfer Learning ◽  
Imaging Technique ◽  
Real Data ◽  
Fine Tuning ◽  
X Ray ◽  
Stenosis Detection ◽  
Artery Disease ◽  
Sensitivity Specificity

Coronary artery disease is the most frequent type of heart disease caused by an abnormal narrowing of coronary arteries, also called stenosis or atherosclerosis. It is also the leading cause of death globally. Currently, X-ray Coronary Angiography (XCA) remains the gold-standard imaging technique for medical diagnosis of stenosis and other related conditions. This paper presents a new method for the automatic detection of coronary artery stenosis in XCA images, employing a pre-trained (VGG16, ResNet50, and Inception-v3) Convolutional Neural Network (CNN) via Transfer Learning. The method is based on a network-cut and fine-tuning approach. The optimal cut and fine-tuned layers were selected following 20 different configurations for each network. The three networks were fine-tuned using three strategies: only real data, only artificial data, and artificial with real data. The synthetic dataset consists of 10,000 images (80% for training, 20% for validation) produced by a generative model. These different configurations were analyzed and compared using a real dataset of 250 real XCA images (125 for testing and 125 for fine-tuning), regarding their randomly initiated CNNs and a fourth custom CNN, trained as well with artificial and real data. The results showed that pre-trained VGG16, ResNet50, and Inception-v3 cut on an early layer and fine-tuned, overcame the referencing CNNs performance. Specifically, Inception-v3 provided the best stenosis detection with an accuracy of 0.95, a precision of 0.93, sensitivity, specificity, and F1 score of 0.98, 0.92, and 0.95, respectively. Moreover, a class activation map is applied to identify the high attention regions for stenosis detection.

Coronary Artery Disease Screening Using CT Coronary Angiography

2020 ◽  
Vol 91 (10) ◽  
pp. 812-817
Author(s):  
Randy Wang Long Cheong ◽  
Brian See ◽  
Benjamin Boon Chuan Tan ◽  
Choong Hou Koh
Keyword(s):  
Systematic Review ◽  
Coronary Artery Disease ◽  
Coronary Artery ◽  
Coronary Angiography ◽  
Cardiac Risk ◽  
Disease Screening ◽  
Ct Coronary Angiography ◽  
Major Adverse Cardiovascular Events ◽  
Cardiac Events ◽  
Artery Disease

BACKGROUND: The increased utility of CT coronary angiography (CTCA) in cardiovascular screenings of aircrew has led to the increased detection of asymptomatic coronary artery disease (CAD). A systematic review of studies relevant to the interpretation of CTCA for the occupational fitness assessment of high-risk vocations was performed, with findings used to describe the development of a pathway for the aeromedical disposition of military aviators with asymptomatic CAD.METHODS: Medline was searched using the terms CT coronary angiogram and screening and prognosis. The inclusion criteria were restricted to study populations ages > 18 yr, were asymptomatic, were not known to have CAD, had undergone CTCA, and with their associations with major adverse cardiovascular events (MACE) and other relevant cardiac outcomes reported.RESULTS: Included in this systematic review were 10 studies. When compared to subjects with no or nonobstructive CAD, those with obstructive CAD on CTCA had hazard ratios (HR) for cardiac events ranging from 1.42 to 105.48. Comparing subjects with nonobstructive CAD and those without CAD on CTCA, a lower HR of 1.19 for cardiac events was found. The annual event rates of subjects with no CAD on CTCA were extremely low, ranging from 0 to 0.5%.CONCLUSIONS: Based on the findings, we suggest that CTCA should only be performed in aircrew with higher cardiac risk profiles. Those found to have no CAD or minimal CAD (i.e., 25% stenosis) in a non-left main coronary artery on CTCA can be returned to flying duties. All other results should be further evaluated with an invasive angiogram.Cheong RWL, See B, Tan BBC, Koh CH. Coronary artery disease screening using CT coronary angiography. Aerosp Med Hum Perform. 2020; 91(10):812817.

scholarly journals Accuracy of MSCT Coronary Angiography with 64 Row CT Scanner—Facing the Facts

2010 ◽  
Vol 4 ◽  
pp. CMC.S3864 ◽  
Author(s):  
M. Wehrschuetz ◽  
E. Wehrschuetz ◽  
H. Schuchlenz ◽  
G. Schaffler
Keyword(s):  
Coronary Artery Disease ◽  
Coronary Artery ◽  
Coronary Angiography ◽  
Positive Predictive Value ◽  
Diagnostic Accuracy ◽  
Predictive Value ◽  
Diagnostic Potential ◽  
Coronary Artery Stenoses ◽  
Artery Disease ◽  
Msct Angiography

Improvements in multislice computed tomography (MSCT) angiography of the coronary vessels have enabled the minimally invasive detection of coronary artery stenoses, while quantitative coronary angiography (QCA) is the accepted reference standard for evaluation thereof. Sixteen-slice MSCT showed promising diagnostic accuracy in detecting coronary artery stenoses haemodynamically and the subsequent introduction of 64-slice scanners promised excellent and fast results for coronary artery studies. This prompted us to evaluate the diagnostic accuracy, sensitivity, specificity, and the negative und positive predictive value of 64-slice MSCT in the detection of haemodynamically significant coronary artery stenoses. Thirty-seven consecutive subjects with suspected coronary artery disease were evaluated with MSCT angiography and the results compared with QCA. All vessels were considered for the assessment of significant coronary artery stenosis (diameter reduction ≥ 50%). Thirteen patients (35%) were identified as having significant coronary artery stenoses on QCA with 6.3% (35/555) affected segments. None of the coronary segments were excluded from analysis. Overall sensitivity for classifying stenoses of 64-slice MSCT was 69%, specificity was 92%, positive predictive value was 38% and negative predictive value was 98%. The interobserver variability for detection of significant lesions had a κ-value of 0.43. Sixty-four-slice MSCT offers the diagnostic potential to detect coronary artery disease, to quantify haemodynamically significant coronary artery stenoses and to avoid unnecessary invasive coronary artery examinations.

Sign in / Sign up

Export Citation Format

Share Document