A meta analysis and hierarchical classification of HU-based atherosclerotic plaque characterization criteria

Can Hounsfield unit density value accurately predict prelaryngeal invasion in laryngeal carcinoma cases

OBJECTIVE

The Hounsfield unit density value (HUDV) is a relative quantitative measurement of radio density used by radiologists in the interpretation of computed tomography (CT) images. Our aim is to investigate the role of HUDV in evaluating pre-epiglottic space (PES) involvement of laryngeal carcinoma.

METHODS

Seventy-four patients treated for laryngeal carcinoma in our clinic between 2014 and 2019 were included in the study. The invasion status of PES was determined radiologically and pathologically. HUDV was measured with a circular selected region of interest, with a constant size of 10 mm2 for PES. The relationship between patological PES invasion, radiological PES invasion, and HUDV was evaluated.

RESULTS

Measuring HUDV to determine PES invasion (74.3 %) was significantly higher than​​ conventional CT evaluation (59.5 %) (p = 0.001). The agreement coefficient (kappa value) of the conventional CT evaluation and the HUDV regarding PES involvement was 0.673, which was interpreted as 'good'.

CONCLUSION

HUDV could be used as an additional tool in diagnosing pre-epiglottic space invasion in laryngeal cancer.

Automatic Coronary Artery Plaque Quantification and CAD-RADS Prediction Using Mesh Priors

Coronary artery disease (CAD) remains the leading cause of death worldwide. Patients with suspected CAD undergo coronary CT angiography (CCTA) to evaluate the risk of cardiovascular events and determine the treatment. Clinical analysis of coronary arteries in CCTA comprises the identification of atherosclerotic plaque, as well as the grading of any coronary artery stenosis typically obtained through the CAD-Reporting and Data System (CAD-RADS). This requires analysis of the coronary lumen and plaque. While voxel-wise segmentation is a commonly used approach in various segmentation tasks, it does not guarantee topologically plausible shapes. To address this, in this work, we propose to directly infer surface meshes for coronary artery lumen and plaque based on a centerline prior and use it in the downstream task of CAD-RADS scoring. The method is developed and evaluated using a total of 2407 CCTA scans. Our method achieved lesion-wise volume intraclass correlation coefficients of 0.98, 0.79, and 0.85 for calcified, non-calcified, and total plaque volume respectively. Patient-level CAD-RADS categorization was evaluated on a representative hold-out test set of 300 scans, for which the achieved linearly weighted kappa ( κ ) was 0.75. CAD-RADS categorization on the set of 658 scans from another hospital and scanner led to a κ of 0.71. The results demonstrate that direct inference of coronary artery meshes for lumen and plaque is feasible, and allows for the automated prediction of routinely performed CAD-RADS categorization.

In vitro analysis of carotid lesions using a preliminary microwave sensor to detect vulnerable plaques: Correlation with histology, Duplex ultrasound examination, and computed tomography scanner: The Imaging and Microwave Phenotyping Assessment of Carotid stenosis Threat (IMPACT) study

Objective

Progress in best medical treatment have made identification of best candidates for carotid surgery more difficult. New diagnostic modalities could be helpful in this perspective. Microwaves (MWs) can quantify dielectric properties (complex relative permittivity) of biological tissues and MW technology has emerged as a promising field of research for distinguishing abnormal tissues from healthy ones. We here evaluated the ability of a dedicated MW sensor developed in our laboratory to identify vulnerable carotid lesions.

Methods

We included 50 carotid lesions in this study. The plaques were analyzed and classified preoperatively by ultrasound (US) examination, computed tomography angiography and tested postoperatively using a MW sensor. Histopathological analysis was used as a gold standard to separate vulnerable plaques (VPs) from nonvulnerable plaques (NVPs).

Results

VPs were more frequently types 2 or 3 plaques (on US examination), had a greater proportion of low (<60 Hounsfield unit) and moderate (60-130 Hounsfield unit) attenuation components (computed tomography angiography) and displayed higher dielectric constant values (MW) than NVPs, which had an opposite profile. NVPs were more frequently asymptomatic plaques compared with VPs (P = .035). Multivariate analysis showed that US examination and MW identified VPs with a sensitivity of 77% and a specificity of 76% (cutoff value, -0.045; area under the curve, 0.848; P < .0001).

Conclusions

We found that the presence of types 2 to 3 (on US examination) and high dielectric constant plaques in vitro was highly indicative of a VP based on histological analysis. Further studies are needed to determine the potential of MW to identify the most dangerous asymptomatic carotid lesions.

Plaque characteristics on coronary CT angiography associated with the positive findings of fractional flow reserve and instantaneous wave-free ratio

BACKGROUND

Fractional flow reserve (FFR) and instantaneous wave-free ratio (iFR) are useful in determining indications for revascularization of coronary artery disease (CAD). Although the discordance of FFR and iFR was noted in approximately 20%, this cause has not been well established. We investigated patient background and features on coronary CT angiography (CCTA) showing not only FFR- and iFR-positive findings but also discordance between FFR ≤ 0.8 and iFR ≤ 0.89.

METHODS

Subjects were consecutively treated in 83 cases with 105 vessels in which stenosis of 30-90% was detected at one vessel of at least 2 mm or more in the major epicardial vessels and FFR and iFR was performed within subsequent 90 days, among suspected CAD which underwent CCTA. The factors affecting not only FFR- and iFR-positive findings, respectively, but also discordance between FFR and iFR was evaluated using logistic regression analysis on per-patient and per-vessel basis.

RESULTS

FFR- and iFR-positive findings were observed in 42 vessels (40.0%) and 34 vessels (32.3%), respectively. Discordance between FFR ≤ 0.8 and iFR ≤ 0.89 was observed in 22 vessels (21.0%) of 21 patients. In multivariate logistic analysis, LAD (OR 3.55; 95%CI 1.20-11.71; p = 0.0217) and lumen volume/myocardial weight (L/M) ratio (OR 0.93; 0.86-0.99, p = 0.0290) were significant predictors for FFR-positive findings. For iFR-positive findings, LAD (OR 3.86; 95%CI 1.12-13.31; p = 0.0236) was only significant predictor. In FFR ≤ 0.8 and iFR > 0.89 group (15 vessels, 14.3%), positive remodeling (PR) (OR 5.03, 95%CI 1.23-20.48, p = 0.0205) was significant predictor. In FFR > 0.8 and iFR ≤ 0.89 group (7 vessels, 6.7%), there were no significant predictors.

CONCLUSION

On CCTA characteristics, a relevant predictor for FFR-positive findings included low L/M ratio. PR was significant predictor in FFR-positive, iFR-negative patients among those with discordance between the FFR and iFR.

Low-density vulnerable thrombus/plaque volume on preoperative computed tomography predicts for spinal cord ischemia after endovascular repair for thoracic aortic aneurysm

BACKGROUND

Similar to open surgical repair, thoracic endovascular aortic repair (TEVAR) carries a risk of spinal cord ischemia (SCI). However, the generally lower incidence of SCI after TEVAR compared with that after open surgical repair, despite the inability to preserve the intercostal arteries, indicates different pathophysiologic mechanisms with the two procedures. We hypothesized that a microembolism from an aortic mural thrombus is the main cause of SCI. Thus, we evaluated the association between the density of a mural thrombus in the descending thoracic aorta and the development of SCI.

METHODS

A retrospective review of a prospectively assembled database was performed for all patients who had undergone surgery at a single institution from October 2008 to December 2018. Patient demographics and procedure-related variables were collected. The volume and Hounsfield unit (HU) value of mural thrombi in the whole descending thoracic aorta were estimated on preoperative computed tomography using a three-dimensional workstation. Logistic regression analysis was performed to identify the risk factors for SCI development.

RESULTS

Of the 367 patients who had undergone TEVAR during the study period, 155 were excluded because of previous arch surgery (n = 59), previous descending thoracic aortic surgery (n = 6), previous TEVAR (n = 6), unavailability of optimal preoperative computed tomography data (n = 17), double-barreled dissection (n = 40), and other reasons. The mean ± standard deviation age of the remaining 212 patients was 75.8 ± 6.4 years, and 42 (19.8%) were women. Of the 212 patients, 14 (6.6%) developed SCI after TEVAR. The low mean density of the mural thrombus, total thrombus volume, low-density (≥-100 HU but <30 HU) thrombus volume, intermediate-density (≥30 HU but <150 HU) thrombus volume, treatment length, urgent surgery, and baseline dialysis differed significantly between patients with and without SCI. Although subsequent multivariate analysis could not be performed owing to the small number of SCI events, vulnerable low-density thrombus/plaque was a stronger predictor among the aneurysm-related factors of SCI after TEVAR on univariate analysis. Well-known risk factors, such as distal coverage between T8 and L1, left subclavian artery coverage, previous abdominal aortic surgery, and prophylactic spinal drainage, did not show significant differences.

CONCLUSIONS

The results from the present study have demonstrated that among aneurysm-related factors, a lower density mural thrombus/plaque in the descending thoracic aorta is a predictor of SCI development after TEVAR. These results suggest that microembolism is one of the important mechanisms of SCI after TEVAR, which might change the prophylactic strategy.

Machine Learning for Assessment of Coronary Artery Disease in Cardiac CT: A Survey

Cardiac computed tomography (CT) allows rapid visualization of the heart and coronary arteries with high spatial resolution. However, analysis of cardiac CT scans for manifestation of coronary artery disease is time-consuming and challenging. Machine learning (ML) approaches have the potential to address these challenges with high accuracy and consistent performance. In this mini review, we present a survey of the literature on ML-based analysis of coronary artery disease in cardiac CT. We summarize ML methods for detection and characterization of atherosclerotic plaque as well as anatomically and functionally significant coronary artery stenosis.

Machine-learning integration of CT histogram analysis to evaluate the composition of atherosclerotic plaques: Validation with IB-IVUS

BACKGROUND

To determine whether machine learning with histogram analysis of coronary CT angiography (CCTA) yields higher diagnostic performance for coronary plaque characterization than the conventional cut-off method using the median CT number.

METHODS

We included 78 patients with 78 coronary plaques who had undergone CCTA and integrated backscatter intravascular ultrasound (IB-IVUS) studies. IB-IVUS diagnosed 32 as fibrous- and 46 as fatty or fibro-fatty plaques. We recorded the coronary CT number and 7 histogram parameters (minimum and mean value, standard deviation (SD), maximum value, skewness, kurtosis, and entropy) of the plaque CT number. We also evaluated the importance of each feature using the Gini index which rates the importance of individual features. For calculations we used XGBoost. Using 5-fold cross validation of the plaque CT number, the area under the receiver operating characteristic curve of the machine learning- (extreme gradient boosting) and the conventional cut-off method was compared.

RESULTS

The median CT number was 56.38 Hounsfield units (HU, 8.00-95.90) for fibrous- and 1.15 HU (-35.8-113.30) for fatty- or fibro-fatty plaques. The calculated optimal threshold for the plaque CT number was 36.1 ± 2.8 HU. The highest Gini index was the coronary CT number (0.19) followed by the minimum value (0.17), kurtosis (0.17), entropy (0.14), skewness (0.11), the mean value (0.11), the standard deviation (0.06), and the maximum value (0.05), and energy (0.00). By validation analysis, the machine learning-yielded a significantly higher area under the curve than the conventional method (area under the curve 0.92 and 95%, confidence interval 0.86-0.92 vs 0.83 and 0.75-0.92, p = 0.001).

CONCLUSION

The machine learning-was superior the conventional cut-off method for coronary plaque characterization using the plaque CT number on CCTA images.

Incremental Prognostic Value of Coronary Computed Tomography Angiography: High-Risk Plaque Characteristics in Asymptomatic Patients

AIM

Coronary computed tomography angiography (CCTA) findings of positive remodeling (index ＞1.1) and low-attenuation plaque (＜30 Hounsfield units) are recognized as CT-verified high-risk plaque (CT-HRP). Therefore, we investigated the incremental prognostic value of evaluation of plaque characteristics using CCTA in asymptomatic patients.

METHODS

Overall, 495 consecutive patients without any known coronary artery disease who underwent CCTA were included in this study. Patients who underwent revascularization within 30 days of CCTA or had scans with poor image quality were excluded. Clinical follow-up data (716.5±262.6 days) were available for 339 patients, who were analyzed for the current study. Framingham risk score (FRS), coronary artery calcium score (CACS), and CT-HRP were investigated as predictors of cardiac events by multivariable analysis using Cox proportional hazard model. Improvement of predictive accuracy by including CT findings was evaluated from reclassification [net reclassification indices (NRI) and integrated discrimination improvement (IDI)] standpoints.

RESULTS

During the follow-up period, 9 cardiac events (cardiac death: 0, nonfatal myocardial infarction: 2, hospitalization for unstable or progressive angina: 7) occurred. Multivariate Cox proportional hazard analysis demonstrated that CACS (HR, 13.23; 95% CI, 1.62-107.78, p＜0.0164) and CT-HRP (HR, 11.27; 95% CI, 1.24-102.12, p＜0.0321) were the independent predictors of cardiac events. NRI was 0.9556 (p＜0.0007) and IDI was 0.2582 (p＜0.0203), and the diagnostic performance improved by CT-HRP added to the combination of CACS and FRS.

CONCLUSION

Although the cardiac event rate was low, the evaluation of CCTA plaque characteristics may provide incremental prognostic value to CACS in asymptomatic patients.

BMP4 enhances foam cell formation by BMPR-2/Smad1/5/8 signaling

Atherosclerosis and its complications are characterized by lipid-laden foam cell formation. Recently, an obvious up-regulation of BMP4 was observed in atherosclerotic plaque, however, its function and the underlying mechanism remains unknown. In our study, BMP4 pretreatment induced macrophage foam cell formation. Furthermore, a dramatic increase in the ratio of cholesteryl ester (CE) to total cholesterol (TC) was observed in BMP4-treated macrophages, accompanied by the reduction of cholesterol outflow. Importantly, BMP4 stimulation inhibited the expression levels of the two most important cellular cholesterol transporters ABCA1 and ABCG1, indicating that BMP4 may induce formation of foam cells by attenuating transporters expression. Further mechanism analysis showed that BMPR-2, one of the BMP4 receptors, was significantly increased in BMP4 treated macrophage foam cells. That blocking its expression using specific siRNA significantly increased ABCA1 and ABCG1 levels. Additionally, BMP4 treatment triggered the activation of Smad1/5/8 pathway by BMPR-2 signaling. After blocking the Smad1/5/8 with its inhibitor, ABCA1 and ABCG1 expression levels were up-regulated significantly, suggesting that BMP4 inhibited the expression of ABCA1 and ABCG1 through the BMPR-2/Smad1/2/8 signaling pathway. Therefore, our results will provide a new insight about how BMP4 accelerate the progressio of atherosclerosis, and it may become a potential target against atherosclerosis and its complications.