Fully Automated CT Quantification of Epicardial Adipose Tissue by Deep Learning: A Multicenter Study

Abstract

To evaluate the performance of deep learning for robust and fully automated quantification of epicardial adipose tissue (EAT) from multicenter cardiac CT data. Materials and Methods:In this multicenter study, a convolutional neural network approach was trained to quantify EAT on non-contrast material-enhanced calcium-scoring CT scans from multiple cohorts, scanners, and protocols (n = 850). Deep learning performance was compared with the performance of three expert readers and with interobserver variability in a subset of 141 scans. The deep learning algorithm was incorporated into research software. Automated EAT progression was compared with expert measurements for 70 patients with baseline and follow-up scans.Results: Automated quantification was performed in a mean (6 standard deviation) time of 1.57 seconds 6 0.49, compared with 15 minutes for experts. Deep learning provided high agreement with expert manual quantification for all scans (R = 0.974; P , .001), with no significant bias (0.53 cm 3 ; P = .13). Manual EAT volumes measured by two experienced readers were highly correlated (R = 0.984; P , .001) but with a bias of 4.35 cm 3 (P , .001). Deep learning quantifications were highly correlated with the measurements of both experts (R = 0.973 and R = 0.979; P , .001), with significant bias for reader 1 (5.11 cm 3 ; P , .001) but not for reader 2 (0.88 cm 3 ; P = .26). EAT progression by deep learning correlated strongly with manual EAT progression (R = 0.905; P , .001) in 70 patients, with no significant bias (0.64 cm 3 ; P = .43), and was related to an increased noncalcified plaque burden quantified from coronary CT angiography (5.7% vs 1.8%; P = .026). Conclusion:Deep learning allows rapid, robust, and fully automated quantification of EAT from calcium scoring CT. It performs as well as an expert reader and can be implemented for routine cardiovascular risk assessment.

Assessing the Role of Pericardial Fat as a Biomarker Connected to Coronary Calcification—A Deep Learning Based Approach Using Fully Automated Body Composition Analysis

“…While manual chest CT annotation is not feasible in clinical routine, our BCA software predicted full tissue quantification in five to ten seconds, making it more resource-conserving and therefore applicable in clinical practice. The system is able to provide various other data on tissue quantification and their relations to each other (e.g., see Section 2.3 ), contrary to other DL-based approaches that focus on single biomarkers, e.g., EAT [ 16 , 21 ].…”

“…Artificial Intelligence is an emerging tool for biomarker extraction and precisely quantifying biomarkers in large-scale study cohorts [ 19 , 20 , 21 ]. Some deep learning-based (DL) approaches to EAT have already been demonstrated [ 16 , 21 ]. While CT scans are performed for various clinical indications, potentially valuable biometric data often goes unused.…”

Epicardial adipose tissue is associated with extent of pneumonia and adverse outcomes in patients with COVID-19

“…EAT was defined as all adipose tissue enclosed by the visceral pericardium [32,33]. EAT volume and attenuation were quantified using a deep learning algorithm incorporated into research software (QFAT version 2.0; Cedars-Sinai Medical Center).…”

“…EAT volume and attenuation were quantified using a deep learning algorithm incorporated into research software (QFAT version 2.0; Cedars-Sinai Medical Center). The development and validation of this automated method have been described previously [32,33]. Briefly, the limits of the heart were automatically defined as the pulmonary artery bifurcation (superior limit) to the posterior descending artery (inferior limit) and pericardial contours traced by the algorithm.…”

Automatic Epicardial Fat Segmentation and Quantification of CT Scans Using Dual U-Nets With a Morphological Processing Layer

“…Commandeur et al first applied a convolutional neural network to obtain epicardial-paracardial masks of the heart. Then, they detected the contour of the pericardium and acquired the final masks of the epicardial and thoracic fat [23][24]. Their method attempted to segment two kinds of cardiac fats simultaneously, which can cause some errors in the segmentation.…”