Abstract:Background: Pathologies of the thoracic aorta are associated with chronic cardiovascular disease and can be of life-threatening nature. Understanding determinants of thoracic aortic morphology is crucial for precise diagnostics and preventive and therapeutic approaches. This study aimed to automatically characterize ascending aortic morphology based on 3D non-contrast-enhanced magnetic resonance angiography (NE-MRA) data from the large epidemiological cross-sectional German National Cohort (NAKO) and to investigate possible determinants of mid-ascending aortic diameter (mid-AAoD). Methods: Deep learning was used to automatically segment the thoracic aorta and extract ascending aortic length, volume, and diameter from 25,073 NE-MRAs. Descriptive statistics, correlation analyses, and multivariable regression were used to investigate statistical relationships between mid-AAoD and demographic factors, hypertension, diabetes, alcohol, and tobacco consumption. Additionally, automated causal discovery analysis using the Peter-Clark algorithm was performed to identify possible causal interactions. Results: Males exhibited significantly larger mid-AAoD than females (M: 35.5±4.8 mm, F: 33.3±4.5 mm). Age and body surface area (BSA) were positively correlated with mid-AAoD. Hypertensive and diabetic subjects showed higher mid-AAoD. Hypertension was linked to higher mid-AAoD regardless of age and BSA, while diabetes and mid-AAoD were uncorrelated across age-stratified subgroups. Daily alcohol consumption and smoking history exceeding 16.5 pack-years exhibited highest mid-AAoD. Causal analysis revealed that age, BSA, hypertension, and alcohol consumption are possibly causally related to mid-AAoD, while diabetes and smoking are likely spuriously correlated. Conclusions: Mid-AAoD varies significantly within the unique large-scale NAKO population depending on demographic factors, individual health, and lifestyle. This work provides a proof-of-concept for automated causal analysis which can help disentangle observed correlations and identify potential causal determinants of ascending aortic morphology.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to understand the morphological changes of the thoracic aorta and its determinants. Specifically, the study aims to characterize the morphology of the ascending aorta by automatically analyzing 25,073 non - contrast - enhanced magnetic resonance angiography (NE - MRA) data in the German national cohort (NAKO) using deep - learning techniques, and to explore the factors that may affect the mid - segment ascending aortic diameter (mid - AAoD). These factors include demographic characteristics (such as gender, age), health conditions (such as hypertension, diabetes), and lifestyle (such as alcohol and tobacco consumption). ### Research Background The thoracic aorta is an important part of the cardiovascular system and is responsible for transporting blood from the heart to the whole body. Understanding the pathological changes of the thoracic aorta is crucial for accurate diagnosis and preventive treatment. These pathological changes may include progressive aneurysm expansion and life - threatening aortic dissection. In addition to genetic factors (such as Marfan syndrome and Ehlers - Danlos syndrome), male gender, age, body surface area (BSA), hypertension, and alcohol and tobacco consumption are also considered potential risk factors for thoracic aortic dilation. However, in the general population, the distribution and determinants of the thoracic aorta have not been fully studied, mainly due to the lack of large - scale population - based studies and sufficient aortic imaging data. ### Research Objectives 1. **Automatically characterize the morphology of the ascending aorta**: Use deep - learning techniques to perform 3D shape analysis on the ascending aorta and extract its length, volume, and diameter. 2. **Descriptive statistical analysis**: Perform descriptive statistical analysis on the aortic phenotype in the general population. 3. **Causal analysis**: Based on a large - scale cross - sectional study, investigate the causal determinants of the mid - segment ascending aortic diameter (mid - AAoD). ### Methods - **Study population**: 25,073 participants were selected from the NAKO database, and data with unqualified image quality, missing data, or unreasonable data were excluded. - **Shape analysis**: Use a deep - learning model to segment NE - MRA images and detect six key points to automatically determine the length, volume, and diameter of the ascending aorta. - **Statistical and causal analysis**: Use univariate linear regression, stratified subgroup analysis, and multivariate Lasso regression to evaluate the relationship between mid - AAoD and observed variables. In addition, use the Peter - Clark algorithm for causal analysis to identify possible causal relationships. ### Main Findings - **Gender differences**: The mid - AAoD of male participants was significantly larger than that of female participants. - **Influence of age and BSA**: Age and BSA were positively correlated with mid - AAoD. - **Health conditions**: The mid - AAoD of patients with hypertension and diabetes was higher, but the correlation between diabetes and mid - AAoD was inconsistent in different age groups. - **Lifestyle**: Daily alcohol consumption and a smoking history of more than 16.5 pack - years were associated with a higher mid - AAoD. - **Causal analysis**: Age, BSA, hypertension, and alcohol consumption may be causal determinants of mid - AAoD, while diabetes and smoking may be spurious correlations. ### Conclusions This study provides a detailed description of the morphology of the ascending aorta and its determinants through the automated analysis of large - scale cross - sectional data, and identifies possible causal relationships through causal analysis methods, which helps to more accurately understand the pathological changes of the thoracic aorta. This provides an important scientific basis for future diagnosis and treatment strategies.

Determinants of ascending aortic morphology: Cross-sectional deep learning-based analysis on 25,073 non-contrast-enhanced MRI of NAKO

Deep learning enables genetic analysis of the human thoracic aorta

Deep Learning‐Based Analysis of Aortic Morphology From Three‐Dimensional MRI

A Cohort Longitudinal Study Identifies Morphology and Hemodynamics Predictors of Abdominal Aortic Aneurysm Growth

Spatially Distinct Genetic Determinants of Aortic Dimensions Influence Risks of Aneurysm and Stenosis

Thoracic Aortic 3-Dimensional Geometry: Effects of Aging and Genetic Determinants

3D aortic morphology and stiffness in MRI using semi-automated cylindrical active surface provides optimized description of the vascular effects of aging and hypertension

Genome-wide associations of aortic distensibility suggest causality for aortic aneurysms and brain white matter hyperintensities

Association of aortic diameters and mortality in the general population—an MRI-based study

Four-dimensional flow cardiovascular magnetic resonance aortic cross-sectional pressure changes and their associations with flow patterns in health and ascending thoracic aortic aneurysm

Automated quantitative 3D analysis of aorta size, morphology, and mural calcification distributions

The Genetic Determinants of Aortic Distention

Flow patterns in ascending aortic aneurysms: Determining the role of hypertension using phase contrast magnetic resonance and computational fluid dynamics

Fully automated aortic diameter measurements on SPECT/CT attenuation maps in patients undergoing myocardial perfusion imaging

Predicting the risk for acute type B aortic dissection in hypertensive patients using anatomic variables

A statistical shape modelling framework to extract 3D shape biomarkers from medical imaging data: assessing arch morphology of repaired coarctation of the aorta

Potential influencing factors of aortic diameter at specific segments in population with cardiovascular risk

Aortic elongation part II: the risk of acute type A aortic dissection

Deep Learning Based Automatic Segmentation of the Thoracic Aorta from Chest Computed Tomography in Healthy Korean Adults

Pathophysiology of the ascending aorta: Impact of dilation and valve phenotype on large-scale blood flow coherence detected by 4D flow MRI

Impact of training data composition on the generalizability of CNN aortic cross section segmentation in 4D Flow MRI