Kernel Conversion for Robust Quantitative Measurements of Archived Chest Computed Tomography Using Deep Learning-Based Image-to-Image Translation

Chest computed tomography (CT) is used to screen for lung cancer and evaluate pulmonary and extra-pulmonary abnormalities such as emphysema and coronary artery calcification, particularly in smokers. In real-world practice, lung abnormalities are visually assessed using high-contrast thin-slice images which are generated from raw scan data using sharp reconstruction kernels with the sacrifice of increased image noise. In contrast, accurate CT quantification requires low-contrast thin-slice images with low noise, which are generated using soft reconstruction kernels. However, only sharp-kernel thin-slice images are archived in many medical facilities due to limited data storage space. This study aimed to establish deep neural network (DNN) models to convert sharp-kernel images to soft-kernel-like images with a final goal to reuse historical chest CT images for robust quantitative measurements, particularly in completed previous longitudinal studies. By using pairs of sharp-kernel (input) and soft-kernel (ground-truth) images from 30 patients with chronic obstructive pulmonary disease (COPD), DNN models were trained. Then, the accuracy of kernel conversion based on the established DNN models was evaluated using CT from independent 30 smokers with and without COPD. Consequently, differences in CT values between new images converted from sharp-kernel images using the established DNN models and ground-truth soft-kernel images were comparable with the inter-scans variability derived from repeated phantom scans (6 times), showing that the conversion error was the same level as the measurement error of the CT device. Moreover, the Dice coefficients to quantify the similarity between low attenuation voxels on given images and the ground-truth soft-kernel images were significantly higher on the DNN-converted images than the Gaussian-filtered, median-filtered, and sharp-kernel images (p < 0.001). There were good agreements in quantitative measurements of emphysema, intramuscular adipose tissue, and coronary artery calcification between the converted and the ground-truth soft-kernel images. These findings demonstrate the validity of the new DNN model for kernel conversion and the clinical applicability of soft-kernel-like images converted from archived sharp-kernel images in previous clinical studies. The presented method to evaluate the validity of the established DNN model using repeated scans of phantom could be applied to various deep learning-based image conversions for robust quantitative evaluation.

Triglyceride to High-Density Lipoprotein Ratio can predict coronary artery calcification

Objectives: We assessed the TG/HDL-C ratio as a predictor for the presence of coronary artery calcifications (CACs). Methods: We collected demographic characteristics (age and gender), physical examination (height, weight, BMI, SBP, DBP), comorbidities, medication use, and laboratory variables Triglyceride to High-Density Lipoprotein (TG, HDL-C, TG/HDL-C, UA, TBG, 25-OH-VitD3); and we used coronary angiography to determine the presence of CACs. We performed univariate and multivariate analyses to evaluate the correlation between the TG/HDL-C ratio and CACs and established a predictive model. Results: CAC was present in 121 patients (25.80%). The levels of TG and TG/HDL-C ratio in the CAC group were higher than those in the non-CAC group, while the level of HDL-C in the CAC group was lower than that in the non-CAC group. The univariate analysis showed that the TG/HDL-C ratio was associated with CAC (OR, 0.021; 95% CI, 0.008 to 0.052; P<0.001), and the multivariate analysis indicated that the ratio was an independent risk factor for CAC (OR, 4.088; 95% CI, 2.787-5.996; P<0.001). Using the ratio to establish a prediction model, the area under the ROC curve was 0.814 (95% CI, 0.775-0.853; P<0.001), suggesting that the TG/HDL-C ratio has a high diagnostic efficiency. The diagnostic threshold was 1.037, and the corresponding sensitivity and specificity were 89.3% and 60.5%, respectively. Conclusion: The Triglyceride to High-Density Lipoprotein TG/HDL-C ratio is an independent risk factor for CAC with good diagnostic efficacy. Abbreviations: TG: Triglycerides, HDL-C: High-Density Lipoprotein, CAC: Coronary Artery Calcifications, BMI: Body Mass Index, SBP: Systolic Blood Pressure, DBP: Diastolic Blood Pressure, UA: Uric Acid, FBG: Fasting Blood Glucose, 25-OH-VitD3: 25-Hydroxyvitamin D3, ACEI: Angiotensin-Converting Enzyme Inhibitors, ARB: Angiotensin Receptor Blockers, CCB: Calcium Channel Blockers, ARNI: Angiotensin Receptor-Neprilysin Inhibitor, CAG: Coronary Angiography, AUCROC: Area Under the Receiver Operating Curve. doi: https://doi.org/10.12669/pjms.38.3.5290 How to cite this:Wang B, Hua J, Ma L. Triglyceride to High-Density Lipoprotein Ratio can predict coronary artery calcification. Pak J Med Sci. 2022;38(3):---------. doi: https://doi.org/10.12669/pjms.38.3.5290 This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Association of Trabecular Bone Score-Adjusted Fracture Risk Assessment Tool with Coronary Artery Calcification in Women

The trabecular bone score (TBS) was found to be significantly associated with moderate coronary artery calcification (CAC). The aim of this study was to further explore the association between TBS-adjusted Fracture Risk Assessment Tool (FRAX) and CAC score in women. The electronic medical record database of a regional teaching hospital in southern Taiwan yielded women who received both coronary computed tomography and bone mineral density (BMD) measurement during their general health examination. Health history, anthropomorphic measurements, laboratory results, BMD, and T-scores were obtained. TBS values were calculated from database spine dual-energy X-ray absorptiometry files. Linear regression analyses tested the association between CAC score and 10-year probability of major osteoporotic fracture (MOF) and hip fracture (HF) determined by TBS-adjusted FRAX. Of the 116 women (mean age 55.8 years) studied, 24.1% had osteoporosis. Simple linear regression showed a significant association of CAC score with an increase in MOF and HF risk as measured by TBS-adjusted FRAX. In multiple linear regression adjusted for potential confounders, CAC score remained significantly associated with TBS-adjusted FRAX for right MOF (p = 0.002), left MOF (p = 0 006), right HF (p = 0.005), and left HF (p = 0.015). In conclusion, clinicians should be vigilant to the potential increased risk of coronary events among women with increased TBS-adjusted FRAX for MOF and HF.

Clinical and radiological characteristics of acute pulmonary embolus in relation to 28-day and 6-month mortality

Background Patients with acute pulmonary embolism (PE) exhibit a wide spectrum of clinical and laboratory features when presenting to hospital and pathophysiologic mechanisms differentiating low-risk and high-risk PE are poorly understood. Objectives To investigate the prognostic value of clinical, laboratory and radiological information that is available within routine tests undertaken for patients with acute PE. Methods Electronic patient records (EPR) of patients who underwent Computed Tomography Pulmonary Angiogram (CTPA) scan for the investigation of acute PE during 6-month period (01.01.2016–30.06.2016) were examined. Data was gathered from EPR for patients that met inclusion criteria and all CTPA scans were re-evaluated. Biochemical thresholds of low-grade and high-grade inflammation, serum CRP >10mg/L and >150mg/L and serum albumin concentrations <35g/L and <25 g/L, were combined in the Glasgow Prognostic Score (GPS) and peri-operative Glasgow Prognostic Score (poGPS) respectively. Neutrophil Lymphocyte ratio (NLR) was also calculated. Pulmonary Embolus Severity Index score was calculated. Results Of the total CTPA reports (n = 2129) examined, 245 patients were eligible for inclusion. Of these, 20 (8%) patients had died at 28-days and 43 (18%) at 6-months. Of the 197 non-cancer related presentations, 28-day and 6-month mortality were 3% and 8% respectively. Of the 48 cancer related presentations, 28-day and 6-month mortality were 29% and 58% respectively. On univariate analysis, age ≥65 years (p<0.01), PESI score ≥100(p = <0.001), NLR ≥3(p<0.001) and Coronary Artery Calcification (CAC) score ≥ 6 (p<0.001) were associated with higher 28-day and 6-month mortality. PESI score ≥100 (OR 5.2, 95% CI: 1.1, 24.2, P <0.05), poGPS ≥1 (OR 2.5, 95% CI: 1.2–5.0, P = 0.01) and NLR ≥3 (OR 3.7, 95% CI: 1.0–3.4, P <0.05) remained independently associated with 28-day mortality. On multivariate binary logistic regression analysis of factors associated with 6-month mortality, PESI score ≥100 (OR 6.2, 95% CI: 2.3–17.0, p<0.001) and coronary artery calcification score ≥6 (OR 2.3, 95% CI: 1.1–4.8, p = 0.030) remained independently associated with death at 6-months. When patients who had an underlying cancer diagnosis were excluded from the analysis only GPS≥1 remained independently associated with 6-month mortality (OR 5.0, 95% CI 1.2–22.0, p<0.05). Conclusion PESI score >100, poGPS≥1, NLR ≥3 and CAC score ≥6 were associated with 28-day and 6-month mortality. PESI score ≥100, poGPS≥1 and NLR ≥3 remained independently associated with 28-day mortality. PESI score ≥100 and CAC score ≥6 remained independently associated with 6-month mortality. When patients with underlying cancer were excluded from the analysis, GPS≥1 remained independently associated with 6-month mortality. The role of the systemic inflammatory response (SIR) in determining treatment and prognosis requires further study. Routine reporting of CAC scores in CTPA scans for acute PE may have a role in aiding clinical decision-making regarding treatment and prognosis.

Circulating miRNA-29b and Sclerostin Levels Correlate with Coronary Artery Calcification and Cardiovascular Events in Maintenance Hemodialysis Patients

Objective. Coronary artery calcification (CAC) is a common complication in end-stage renal disease (ESRD) patients undergoing maintenance hemodialysis (MHD), and the extent of CAC is a predominant predictor of cardiovascular outcomes in MHD patients. In this study, we sought to uncover the relationship between circulating miRNA-29b, sclerostin levels, CAC, and cardiovascular events (CVEs) in MHD patients. Methods. This study recruited patients receiving MHD for at least three months in the Hainan General Hospital between January 2016 and June 2019, and all patients were followed up 24 months for CVEs. The serum level of sclerostin was determined by enzyme-linked immunosorbent assay (ELISA) and miRNA-29b expression by real-time qPCR (RT-qPCR). All patients received cardiac CT scans to evaluate CAC, and CAC scores were expressed in Agatston units. The MHD patients with CACs <100 were arranged into the CAC (<100) group, those with 100–400 CACs into the CAC (100–400) group, and those with CACs >400 into the CAC (>400) group. Net reclassification index (NRI) and integrated discrimination index (IDI) were calculated to assess the predictive performance of serum sclerostin level for the occurrence of CVEs. Results. Compared with the CAC (<100) group, the CAC (>400) group had higher proportions of older patients, hypertension and diabetes mellitus patients, longer dialysis duration, higher mean arterial pressure (MAP), higher levels of high-sensitivity C-reactive protein (hs-CRP), alkaline phosphatase (ALP), and phosphate ( P < 0.05 ). It was found that the CAC (100–400) and CAC (>400) groups exhibited higher serum levels of sclerostin but lower levels of miRNA-29b than the CAC (<100) group ( P < 0.05 ) and the CAC (>400) group had a higher level of sclerostin and a lower level of miRNA-29b than the CAC (100–400) group ( P < 0.05 ). The circulating level of miRNA-29b was negatively correlated with the serum level of sclerostin in MHD patients (r = −0.329, P < 0.01 ). The multivariate logistic regression analysis showed that hs-CRP, phosphate, sclerostin, and miRNA-29b were independent risk factors for CAC in MHD patients ( P < 0.05 , Table 2). ROC for prediction of CAC by sclerostin yielded 0.773 AUC with 95% CI 0.683–0.864 ( P < 0.01 ). As depicted by Kaplan–Meier curves of CVE incidence in MHD patients according to median sclerostin (491.88 pg/mL) and median miRNA-29b (Ct = 25.15), we found that serum levels of sclerostin and miRNA-29b were correlated with the incidence of CVEs in MHD patients. When a new model was used to predict the incidence of CVEs, NRI 95% CI was 0.60 (0.16–1.03) ( P < 0.05 ) and IDI 95% CI was 0.002 (−0.014 to 0.025) ( P < 0.05 ), suggesting that sclerostin added into the old model could improve the prediction of the incidence of CVEs. Conclusions. These data suggest that circulating miRNA-29b and sclerostin levels are correlated with CAC and incidence of CVEs in MHD patients. Higher sclerostin and lower miRNA-29b may serve as independent risk factors for the incidence of CVEs in MHD patients.