Deep learning-based plaque quantification from coronary computed tomography angiography: external validation and comparison with intravascular ultrasound

Background Atherosclerotic plaque quantification from coronary computed tomography angiography (CTA) enables accurate assessment of coronary artery disease burden, progression, and prognosis. However, quantitative plaque analysis is time-consuming and requires high expertise. We sought to develop and externally validate an artificial intelligence (AI)-based deep learning (DL) approach for CTA-derived measures of plaque volume and stenosis severity. We compared the performance of DL to expert readers and the gold standard of intravascular ultrasound (IVUS). Methods This was a multicenter study of patients undergoing coronary CTA at 11 sites, with software-based quantitative plaque measurements performed at a per-lesion level by expert readers. AI-based plaque analysis was performed by a DL novel convolutional neural network which automatically segmented the coronary artery wall, lumen, and plaque for the computation of plaque volume and stenosis severity. Using expert measurements as ground truth, the DL algorithm was trained on 887 patients (4,686 lesions). Thereafter, the algorithm was applied to an independent test set of 221 patients (1,234 lesions), which included an external validation cohort of 171 patients from the SCOT-HEART (Scottish Computed Tomography of the Heart) trial as well as 50 patients who underwent IVUS within one month of CTA. We report the performance of AI-based plaque analysis in the independent test set. Results Within the external validation cohort, there was excellent agreement between DL and expert reader measurements of total plaque volume (intraclass correlation coefficient [ICC] 0.876), noncalcified plaque volume (ICC 0.869), and percent diameter stenosis (ICC 0.850; all p<0.001). When compared with IVUS, there was excellent agreement for DL total plaque volume (ICC 0.945), total plaque burden (ICC 0.853), minimal luminal area (ICC 0.864), and percent area stenosis (ICC 0.805; all p<0.001); with strong correlation between DL and IVUS for total plaque volume (r=0.915; p<0.001; Figure). The average DL plaque analysis time was 20 seconds per patient, compared with 25–30 minutes taken by experts. Conclusions AI-based plaque quantification from coronary CTA using an externally validated DL approach enables rapid measurements of plaque volume and stenosis severity in close agreement with expert readers and IVUS. FUNDunding Acknowledgement Type of funding sources: Public grant(s) – National budget only. Main funding source(s): National Heart, Lung, and Blood Institute, United States

Vesseg: An Open-Source Tool for Deep Learning-Based Atherosclerotic Plaque Quantification in Histopathology Image

Objective: Manual plaque segmentation in microscopy images is a time-consuming process in atherosclerosis research and potentially subject to unacceptable user-to-user variability and observer bias. We address this by releasing Vesseg a tool that includes state-of-the-art deep learning models for atherosclerotic plaque segmentation. Approach and Results: Vesseg is a containerized, extensible, open-source, and user-oriented tool. It includes 2 models, trained and tested on 1089 hematoxylin-eosin stained mouse model atherosclerotic brachiocephalic artery sections. The models were compared to 3 human raters. Vesseg can be accessed at https://vesseg .online or downloaded. The models show mean Soerensen-Dice scores of 0.91±0.15 for plaque and 0.97±0.08 for lumen pixels. The mean accuracy is 0.98±0.05. Vesseg is already in active use, generating time savings of >10 minutes per slide. Conclusions: Vesseg brings state-of-the-art deep learning methods to atherosclerosis research, providing drastic time savings, while allowing for continuous improvement of models and the underlying pipeline.

Simple and rapid plaque assay for recombinant baculoviruses expressing influenza hemagglutinin

Recombinant baculoviruses (rBVs) have been extensively used to generate virus-like particles, and baculoviruses expressing antigenic proteins have become efficient tools for inducing protective immunity. However, current methods for generating baculoviruses are costly and inefficient. Thus, the development of a simple, rapid, and accurate method of baculovirus titration is critically important. We established a method of plaque assay using an immunostaining method by which plaques can be easily visualized in Sf9 cells under a light microscope. Sf9 cells were infected with recombinant baculoviruses expressing influenza hemagglutinin surface proteins from H1N1 (A/California/04/09) or rH5N1 (A/Vietnam/1203/04). The infected cells were incubated with anti-HA antibody and the plaques were visualized using the chromogen 3′3-diaminobenzidine (DAB). Plaques were observed from days 1 to 6 post-infection, and differences in Sf9 cell seeding densities resulted in variations in the final plaque quantification. Sf9 cells seeded at a concentration of 5.5 × 104 cells/well or 7.5 × 104 cells/well showed the higher plaque titers at days 3, 4, and 5 post-infection than those found at days 1, 2, and 6 post-infection. With 5.5 × 104 cells/well or 7.5 × 104 cells/well of cell concentrations, recombinant baculovirus for rBV-HA (H1N1) showed 6 × 107 pfu/ml of titer and rBVs for rBV-HA (rH5N1) showed 5.4 × 107 pfu/ml of titer. Three days of baculovirus incubation with a certain concentration of Sf9 cells seeded are required for a rapid, simple, and accurate plaque assay, which could significantly contribute to all baculovirus-related studies.

Prognostic value of peri-coronary adipose tissue attenuation and whole vessel and lesion plaque quantification on Coronary Computed Tomography Angiography

Background Peri-coronary adipose tissue attenuation expressed by fat attenuation index (FAI) on coronary CT angiography (CCTA) reflects peri-coronary inflammation and is associated with cardiac mortality. CCTA also provides two-dimensional and three-dimensional quantification of the individual component of atherosclerotic plaque and entire vessel. The atherosclerotic burden or disease extent in entire epicardial coronary arteries provides prognostic information in patients with coronary artery disease. Purpose This study sought to explore the prognostic significance of FAI values and whole vessel and lesion plaque quantification on CCTA in stable patients with intermediate epicardial stenosis evaluated by fractional flow reserve (FFR). Methods A total of 277 patients (277 lesions) with intermediate coronary stenosis who underwent FFR measurement and CCTA were studied. FAI was assessed by the crude analysis of the mean CT attenuation value (−190 to −30 Hounsfield units; higher values indicating inflammation) on CCTA. CT findings including whole vessel and lesion plaque quantification, and target vessel myocardial mass were investigated. Major adverse cardiovascular outcome (MACE) was defined as all cause death, cardiac death, myocardial infarction, unplanned revascularization, and heart failure requiring admission. Survivals from MACE were assessed. Results The mean FAI and the median FFR values were −71.6 and 0.77, respectively. FFR values were weakly albeit significantly correlated with FAI values. (r=−0.016, P=0.008.) MACE was occurred 43 (15.5%) patients during 5 years F-up. ROC analyses revealed that best cut-off value of FAI to predict MACE was −73.1. Kaplan-Meier analysis revealed that lesions with FAI ≥−73.1 had a significantly higher risk of MACE. (Chi-square 5.5, P=0.019) FFR values and the percutaneous coronary intervention were not predictive of MACE. Multivariate COX proportional hazards regression analysis revealed that age, remodeling index, and lesions with FAI ≥−73.1 were independent predictors of MACE. Conclusion The peri-coronary inflammation evaluated by FAI and CT remodeling index enhances cardiac risk prediction in chronic coronary syndrome patients with intermediate lesions. Non-invasive comprehensive CT assessment may help identify high risk patients of subsequent clinical events and provide enhanced patient management. Funding Acknowledgement Type of funding source: Foundation. Main funding source(s): This study was supported in part by an unrestricted research grant from St. Jude Medical (Abbot Vascular, Santa Clara, CA, USA). The company had no role in study design, conduct, data analysis or manuscript preparation.