CT attenuation values of blood and myocardium: rationale for accurate coronary artery calcifications detection with multi-detector CT

Artificial intelligence-based automated left ventricular mass quantification from non-contrast cardiac CT scans: correlation with contrast CT and cardiac MRI

Background

Non-contrast CT scans are not used for evaluating left ventricle myocardial mass (LV mass), which is typically evaluated with contrast CT or cardiovascular magnetic resonance imaging (MRI). We assessed the feasibility of LV mass estimation from standard, ECG-gated, non-contrast CT using an artificial intelligence (AI) approach and compare it with coronary CT angiography (CTA) and cardiac MRI.

Methods

We enrolled consecutive patients who underwent coronary CTA, which included non-contrast CT calcium scanning and contrast CTA, and cardiac MRI. The median interval between coronary CTA and MRI was 22 days (IQR: 3-76). We utilized an nn-Unet AI model that automatically segmented non-contrast CT structures. AI measurement of LV mass was compared to contrast CTA and MRI.

Results

A total of 316 patients (Age: 57.1±16.7, 56% male) were included. The AI segmentation took on average 22 seconds per case. An excellent correlation was observed between AI and contrast CTA LV mass measures (r=0.84, p<0.001), with no significant differences (136.5±55.3 vs. 139.6±56.9 g, p=0.133). Bland-Altman analysis showed minimal bias of 2.9. When compared to MRI, measured LV mass was higher with AI (136.5±55.3 vs. 127.1±53.1 g, p<0.001). There was an excellent correlation between AI and MRI (r=0.85, p<0.001), with a small bias (-9.4). There were no statistical differences between the correlations of LV mass between contrast CTA and MRI, or AI and MRI.

Conclusions

The AI-based automated estimation of LV mass from non-contrast CT demonstrated excellent correlations and minimal biases when compared to contrast CTA and MRI.

Integrated intensity-based technique for coronary artery calcium mass measurement: A phantom study

BACKGROUND

Agatston scoring, the traditional method for measuring coronary artery calcium, is limited in its ability to accurately quantify low-density calcifications, among other things. The inaccuracy of Agatston scoring is likely due partly to the arbitrary thresholding requirement of Agatston scoring.

PURPOSE

A calcium quantification technique that removes the need for arbitrary thresholding and is more accurate, sensitive, reproducible, and robust is needed. Improvements to calcium scoring will likely improve patient risk stratification and outcome.

METHODS

The integrated Hounsfield technique was adapted for calcium scoring (integrated calcium mass). Integrated calcium mass requires no thresholding and includes all calcium information within an image. This study utilized phantom images acquired by G van Praagh et al., with calcium hydroxyapatite (HA) densities in the range of 200-800 mgHAcm-3 to measure calcium according to integrated calcium mass and Agatston scoring. The calcium mass was known, which allowed for accuracy, reproducibility, sensitivity, and robustness comparisons between integrated calcium mass and Agatston scoring. Multiple CT vendors (Canon, GE, Philips, Siemens) were used during the image acquisition phase, which provided a more robust comparison between the two calcium scoring techniques. Three calcification inserts of different diameters (1, 3, and 5 mm) and different HA densities (200, 400, and 800 mgHAcm-3 ) were placed within the phantom. The effect of motion was also analyzed using a dynamic phantom. All dynamic phantom calcium inserts were 5.0 ± 0.1 mm in diameter with a length of 10.0 ± 0.1 mm. The four different densities were 196 ± 3, 380 ± 2, 408 ± 2, and 800 ± 2 mgHAcm-3 .

RESULTS

Integrated calcium mass was more accurate than Agatston scoring for stationary scans (R M S E I n t e g r a t e d = 2.87 $RMS{E}_{Integrated} = 2.87$ ,R M S E A g a t s o n = 4.07 $RMS{E}_{Agatson} = 4.07$ ) and motion affected scans (R M S E I n t e g r a t e d = 9.70 $RMS{E}_{Integrated} = 9.70$ ,R M S E A g a t s o n = 19.98 $RMS{E}_{Agatson} = 19.98$ ). On average, integrated calcium mass was more reproducible than Agatston scoring for two of the CT vendors. The percentage of false-negative and false-positive calcium scores were lower for integrated calcium mass (15.00%, 0.00%) than Agatston scoring (28.33%, 6.67%). Integrated calcium mass was more robust to changes in scan parameters than Agatston scoring.

CONCLUSIONS

The results of this study indicate that integrated calcium mass is more accurate, reproducible, and sensitive than Agatston scoring on a variety of different CT vendors. The substantial reduction in false-negative scores for integrated calcium mass is likely to improve risk-stratification for patients undergoing calcium scoring and their potential outcome.

Challenging Cases of Aortic Prosthesis Dysfunction, the Importance of Multimodality Imaging, a Case Series

ECG-gated multidetector computed tomography (MDCT) is a promising complementary technique for evaluation of cardiac native and prosthetic structures. MDCT is able to provide a broader coverage with faster scan acquisition times that yield higher spatial and temporal resolution for cardiac structures whose quality may be affected by artifacts on ultrasound. We report a case series about the most challenging complications occurring after prosthetic aortic valve implantation in four patients: pannus, paravalvular leak, prosthesis’ misfolding and subaortic membrane reformation. In all the cases, enhanced MDCT using a retrospective protocol provided accurate 3D morphoanatomic information about cardiac and extracardiac structures, improving and speeding up the correct diagnosis and treatment planning. Integrated imaging, in particular with MDCT, is now the present, and it will increasingly be the future in the assessment of cardiac structural pathology.

Sources of error in bone mineral density estimates from quantitative CT

Bone mineral density (BMD) estimates from quantitative computed tomography (QCT) have proven useful for opportunistic screening of osteoporosis, treatment monitoring, and bone strength measurement. These estimates are subject to bias and variance from a variety of sources related to the imaging equipment, methods applied in the estimation procedure, and the patients themselves. In this article, we review the literature to describe the sources and sizes of error in spine and hip BMD estimates from single-energy QCT that can result from factors related to the scanner, imaging techniques, imaging subject, calibration phantom, and calibration approach. We also describe the baseline variance that can be expected based on repeatability and reproducibility studies. Though reproducible BMD estimates may be achievable with QCT, a thorough understanding of the potential sources of error and their size relative to the diagnostic task is essential to their appropriate and meaningful interpretation.

Rapid Onset Development of Myocardial Calcifications in the Setting of Renal Failure and Sepsis

Myocardial calcifications can arise following damage to myocardial tissue or in the setting of disturbances in the calcium and phosphorus balance. They are associated with a number of cardiac sequelae, as well as higher mortality. Three cases of rapid-onset myocardial calcifications that developed within the course of 5 to 13 weeks in patients who had a history of sepsis and renal failure while undergoing hemodialysis are described. Baseline imaging from several weeks prior without myocardial calcification are shown for each of the three patients, demonstrating the rapid onset of these calcifications. The clinical significance of these findings is discussed. © RSNA, 2021.

Coronary Calcium Detection Based on Improved Deep Residual Network in Mimics

Coronary calcium detection in medicine image processing is a hot research topic. According to the low resolution and complex background in medicine image, an improved coronary calcium detection algorithm based on the Single Shot MultiBox Detector (SSD) in Mimics is proposed in this paper. The algorithm firstly uses the aggregate channel feature model to preprocess the image to obtain the suspected calcium area, which greatly reduces the time of single-frame image detection. The basic network VGG-16 is replaced by Resnet-50, which introduces the identity mapping to solve the problem of reducing the detection accuracy when the number of network layers are increased. Finally, the powerful and flexible two-parameter loss function is used to optimize the training deep network and improve the network model generalization ability. Qualitative and quantitative experiments show that the performance of the proposed detection algorithm exceeds the existing calcium detection algorithms, and the detection efficiency is improved while ensuring the accuracy of calcium detection.