Magnetic Resonance Imaging Of Coronary Arteries Research Articles

Detection of Abnormality in Coronary Artery Magnetic Resonance Imaging using Bit Plane Slicing and Deep Learning.

This paper presents a novel approach for detecting abnormality in coronary arteries using MRI data in RGB images. The study evaluates the test accuracy of the weak classifiers and the test accuracy and F1 score of the strong classifier. The method involves separating the image into information planes, including R, G, and B color space, or bit-planes, and training a VGG-like convolutional neural network model on each plane separately, referred to as a "weak classifier." The classification results of these planes are aggregated using a proposed soft voting method, forming a "strong classifier," with the weights for the aggregation determined by the model's performance on the training set. The results indicate that the strong classifier achieves a test accuracy and F1 score of around 68% to 74% on our private coronary artery dataset. Moreover, by aggregating the top three highest bit-plane levels in a grayscale image, the accuracy is slightly lower than that of the three color spaces but requires a significantly smaller CNN model of nearly 4M parameters. The potential of bit-planes in reducing model storage costs is suggested. This approach holds promise for improving the detection of abnormalities in coronary arteries using MRI data.

Identifying Abnormal Ostial Morphology in Anomalous Aortic Origin of a Coronary Artery

Anomalous origin of a coronary artery from the contralateral sinus of Valsalva is associated with exercise-induced ischemia and sudden death. That is thought to be due to aortic enlargement in patients with an elliptical ostium. We hypothesize that virtual angioscopy can identify abnormal coronary ostial morphology in these patients. We retrospectively analyzed 55 consecutive pediatric coronary artery magnetic resonance imaging studies from January 2006 to January 2010 with the diagnosis of anomalous right (n = 20), or left (n = 7) coronary artery, or normal coronary origins (n = 28). One postmortem heart specimen with anomalous left coronary artery was imaged and analyzed to validate our technique. Virtual angioscopy analysis was used for visualization and measurement of the coronary ostia. Distinct aortic origins of the right and left coronaries were seen in all 55 studies. An elliptical orifice with a longer superior-inferior dimension was seen in all anomalous ostia, in contrast to a circular ostium in all normal origins. That was quantified in anomalous ostia with a long-axis to short-axis ratio of 2.5 ± 0.5 (right) and 2.4 ± 0.5 (left) compared with 1.1 ± 0.2 (right) and 1.0 ± 0.3 (left) in controls (p < 0.001 for right and left ostia comparisons). Ostial morphology was confirmed in all 9 patients who underwent operative repair and in 1 patient at autopsy. Virtual angioscopy identifies abnormal ostial morphology in anomalous coronary artery patients, which is important for characterizing the diagnosis of patients who may be at risk for sudden death.

Positive effect of sublingual nitroglycerin on performance of non-contrast enhanced magnetic resonance coronary angiography

Background Cardiac magnetic resonance imaging (CMR) is increasingly proposed for non-invasive detection of relevant coronary artery disease (CAD), currently using an integrated assessment of myocardial perfusion, viability and function. Magnetic resonance coronary artery imaging (MRCA) is still experimental. Currently, it is not proven if sublingual nitroglycerin (slNitro) should be given to enhance performance of MRCA. Methods 15 volunteers, 9 (60%) of them male, mean age 37.2 years (±11) without known CAD underwent MRCA, without and with slNitro. slNitro dosage was adjusted according to blood pressure (systolic blood pressure (BP) >180 mmHg =>1.6 mg glyceroltrinitrat, BP 140-179 mmHg =>1.2 mg, BP 100-139 mmHg =>0.8 mg, BP 90-100mmHg =>0.4mg). All volunteers were examined in supine position using a GE Signa HDxt 1.5 Tesla scanner, equipped with EchoSpeed gradients, and a dedicated 8-element phased array cardiac coil (GE Healthcare, Milwaukee, Wisconsin). For imaging of the coronary arteries a commercially available whole heart 3D navigator gated multislab steady state free precession sequence (3D HEART, based on 3D FatSat FIESTA) without the administration of contrast medium was employed. This sequence is designed for free breathing cardiac MR angiography, using a navigator echo pulse that detects motion of the diaphragm. We used an abdominal belt in all vol to reduce motion of the diaphragm. Post processing was perfomed with cvi 42 5.0 (Circle, Calgary, Canada). Vessel diameter was measured before and after slNitro. Image quality was graded visually before and after slNitro on a 4-point scale: 1 = non-assessable with severe image artifacts, poor vessel contrast; 2 = assessable with

Improved MR coronary angiography with use of a new rapid clearance blood pool contrast agent in pigs.

To evaluate in an animal model the potential for clinical use of a new rapid clearance blood pool contrast agent to improve free-breathing and breath-hold magnetic resonance (MR) coronary angiography. Free-breathing and breath-hold MR coronary angiography were performed in a pig model (n = 9) (a) without use of a contrast agent; (b) with P792 (Guerbet Research, Aulnay Sous Bois, France), a monodisperse monogadolinated macromolecular compound that acts as a blood pool contrast agent with rapid clearance properties; and (c) with an extravascular gadolinium-based contrast agent. This resulted in six imaging options, which were compared in terms of contrast-to-noise ratio (CNR), signal-to-noise ratio, and vessel length measurements by using the Student t test. Use of P792 improved CNR and visible vessel length significantly with both MR respiratory motion correction approaches, as compared with nonenhanced MR imaging (P <.05). CNR was improved by 76% (from 5.0 to 8.6) with the free-breathing approach and by 34% (from 6.2 to 8.2) with the breath-hold approach. Visible vessel length was increased by 27% (from 79.7 to 99.2 mm) with the free-breathing approach and by 90% (from 48.2 to 86.5 mm) with the breath-hold approach. The P792-enhanced free-breathing approach allowed more distal visualization of the coronary arteries than did the P792-enhanced breath-hold approach (P <.05). Use of the extravascular contrast agent did not improve image quality significantly when compared with that of nonenhanced MR images. Use of P792 improves coronary artery MR imaging in conjunction with free-breathing and breath-hold approaches.

Initial experience with the intravascular contrast agent NC100150-injection (Clariscan) for breath-hold and navigator-gated magnetic resonance coronary artery imaging.

To examine magnetic resonance coronary artery imaging after NC100150-Injection. Breath-hold and navigator-gated images were acquired in five patients. Breath-hold image quality, coronary artery-fat SDNR, and coronary artery SNR improved. Respiratory artifacts due to reduced liver signal intensity degraded navigator-gated image quality. NC100150-Injection improves breath-hold coronary artery imaging. Navigator-gated acquisitions should use techniques that are insensitive to T2* effects.

Three-dimensional coronary artery MR imaging using prospective real-time respiratory navigator and linear phase shift processing: comparison with conventional coronary angiography

Respiratory gating with navigator echo is a recent technique to detect diaphragm position in 3D magnetic resonance (MR) coronary angiography. The purpose of our study was to image proximal coronary arteries and to detect significant stenoses in patients with coronary artery diseases and to compare with contrast enhanced angiography results. Twenty patients with coronary artery diseases who were referred for conventional angiography underwent magnetic resonance angiography (MRA). Three-dimensional gradient echo volumes were acquired using cardiac and respiratory gating and fat suppression. Using reformatted oblique planes and maximum intensity projection technique, visualization coronary segments and detection of significant coronary stenoses were made. Eighty-three coronary segments were analyzed. The sensitivity and specificity were 65% and 93%, respectively. The corresponding positive and negative predictive values were 69% and 91%. This study shows the ability to image correctly coronary arteries and to identify proximal stenoses, but image quality need to be improved for an efficiency detection of coronary artery stenoses in clinical practice.

Magnetic resonance coronary artery imaging for redo coronary operations

Imaging the native coronary vessels using contrast angiography can be difficult in the context of redo coronary operations when native disease is severe. When previous vein grafts undergo aneurysmal dilatation, imaging of the native vessel is restricted by the inability to infuse a sufficient volume of contrast medium through the graft and into the native vessel. We present a case of a patient whose redo coronary artery bypass graft operation was planned on the basis of magnetic resonance imaging of his native coronary arteries and vein graft after unsuccessful coronary angiography.

Magnetic resonance coronary artery imaging.

Coronary artery imaging with magnetic resonance imaging carries the potential for a noninvasive, essentially risk-free screening test for those at risk for coronary heart disease. Many physiologic and anatomic challenges including cardiac and respiratory motion and the small size, tortuosity, and variable flow characteristics of the coronary arteries hamper effort to achieve this goal. This article reviews the efforts of several research groups to surmount these difficulties through the use of 2D and 3D techniques; spin echo, gradient echo, and ultrafast sequences; saturation pulses; and contrast agents. Promising results have been and continue to be reported although no obvious optimal solution has yet been determined.