Automated Coronary Artery Calcium and Quantitative Emphysema in Lung Cancer Screening

Abstract

Purpose: To assess automated coronary artery calcium (CAC) and quantitative emphysema (percentage of low attenuation areas [%LAA]) for predicting mortality and lung cancer (LC) incidence in LC screening. To explore correlations between %LAA, CAC, and forced expiratory value in 1 second (FEV1) and the discriminative ability of %LAA for airflow obstruction. Materials and Methods: Baseline low-dose computed tomography scans of the BioMILD trial were analyzed using an artificial intelligence software. Univariate and multivariate analyses were performed to estimate the predictive value of %LAA and CAC. Harrell C-statistic and time-dependent area under the curve (AUC) were reported for 3 nested models (Modelsurvey: age, sex, pack-years; Modelsurvey-LDCT: Modelsurvey plus %LAA plus CAC; Modelfinal: Modelsurvey-LDCT plus selected confounders). The correlations between %LAA, CAC, and FEV1 and the discriminative ability of %LAA for airflow obstruction were tested using the Pearson correlation coefficient and AUC-receiver operating characteristic curve, respectively. Results: A total of 4098 volunteers were enrolled. %LAA and CAC independently predicted 6-year all-cause (Modelfinal hazard ratio [HR], 1.14 per %LAA interquartile range [IQR] increase [95% CI, 1.05-1.23], 2.13 for CAC ≥400 [95% CI, 1.36-3.28]), noncancer (Modelfinal HR, 1.25 per %LAA IQR increase [95% CI, 1.11-1.37], 3.22 for CAC ≥400 [95%CI, 1.62-6.39]), and cardiovascular (Modelfinal HR, 1.25 per %LAA IQR increase [95% CI, 1.00-1.46], 4.66 for CAC ≥400, [95% CI, 1.80-12.58]) mortality, with an increase in concordance probability in Modelsurvey-LDCT compared with Modelsurvey (P<0.05). No significant association with LC incidence was found after adjustments. Both biomarkers negatively correlated with FEV1 (P<0.01). %LAA identified airflow obstruction with a moderate discriminative ability (AUC, 0.738). Conclusions: Automated CAC and %LAA added prognostic information to age, sex, and pack-years for predicting mortality but not LC incidence in an LC screening setting. Both biomarkers negatively correlated with FEV1, with %LAA enabling the identification of airflow obstruction with moderate discriminative ability.

Cardiac troponin I as predictor for cardiac and other mortality in the German randomized lung cancer screening trial (LUSI)

“…Current research focuses on the complementary use of CT-images for identifying individuals at elevated risk of cardiovascular mortality, on the basis of radiologic ascertainments of emphysema, COPD or coronary artery calcification scores 28 , 29 . A limitation of our study is that for the present analyses we did not dispose of CT-based assessment of emphysema and coronary artery calcification; thus, we could not examine the added value of cTnI measurements for this purpose, compared to radiologic data.…”

Consistency of automated coronary calcium score and extent of emphysema with different CT scanners and radiation dose protocols in lung cancer screening

“…LDCT and ULDCT images were transferred to a dedicated workstation (Alienware Area 51 R6 equipped with Dual NVIDIA GeForce RTX 2080 OC graphics) and analyzed by a fully automated AI software (AVIEW, Coreline Soft). As reported elsewhere [10], CAC was measured with a 3-dimensional U-net architecture-based scoring tool and stratified using three predefined Agatston score categories: 0-99, 100-399, and ≥ 400 [15]. Emphysema was quantified using the percentage of lung volume occupied by voxels with attenuation ≤ −950 Hounsfield Units, HU (percentage of low attenuation areas, %LAA) [16, 17].…”

“…Such a potentially huge number and heterogeneity of imaging data pose the need for a timesaving and reproducible approach for imaging analysis. The radiological literature explored the field of quantitative imaging, leading to the availability of objective parameters for assessing both emphysema and CAC, mostly exploiting Artificial Intelligence (AI) -based systems [10][11][12]. Nonetheless, whether and to what extent an objective and reproducible AI-based quantification of CAC and emphysema is affected by CT equipment and radiation dose is still to be defined.…”

Long-term outcomes of lung cancer screening in males and females

“…LDCT images were transferred to a dedicated graphic station (Alienware Area 51 R6 equipped with Dual NVIDIA GeForce RTX 2080 ◦ C graphics), and CAC was automatically evaluated using commercially available AI-based software (AVIEW, Coreline Soft, Seoul, Korea). CAC was assessed based on the Agatston score and stratified as follows: 0, 1–10, 11–100, 101–400, and > 400 [ 22 , 23 , 27 , 28 ].…”