Simultaneous Multi-Structure Segmentation of the Heart and Peripheral Tissues in Contrast Enhanced Cardiac Computed Tomography Angiography

Longitudinal association of epicardial and thoracic adipose tissues with coronary and cardiac characteristics in psoriasis

Background

s: Psoriasis is a disease of systemic inflammation associated with increased cardiometabolic risk. Epicardial adipose tissue (EAT) and thoracic adipose tissue (TAT) are contributing factors for atherosclerosis and cardiac dysfunction. We strove to assess the longitudinal impact of the EAT and TAT on coronary and cardiac characteristics in psoriasis.

Methods

The study consisted of 301 patients with baseline coronary computed tomography angiography (CTA), of which 139 had four-year follow up scans. EAT and TAT volumes from non-contrast computed tomography scans were quantified by an automated segmentation framework. Coronary plaque characteristics and left ventricular (LV) mass were quantified by CTA.

Results

When stratified by baseline EAT and TAT volume quartiles, a stepwise significant increase in cardiometabolic parameters was observed. EAT and TAT volumes associated with fibro-fatty burden (FFB) (TAT: ρ = 0.394, P < 0.001; EAT: ρ = 0.459, P < 0.001) in adjusted models. Only EAT had a significant four-year time-dependent association with FFB in fully adjusted models (β = 0.307 P = 0.003), whereas only TAT volume associated with myocardial injury in fully adjusted models (TAT: OR = 1.57 95 % CI = (1.00-2.60); EAT: OR = 1.46 95 % CI = (0.91-2.45). Higher quartiles of EAT and TAT had increased LV mass and developed strong correlation (TAT: ρ = 0.370, P < 0.001; EAT: ρ = 0.512, P < 0.001).

Conclusions

Our study is the first to explore how both EAT and TAT volumes associate with increased cardiometabolic risk profile in an inflamed psoriasis cohorts and highlight the need for further studies on its use as a potential prognostic tool for high-risk coronary plaques and cardiac dysfunction.

DeepHeartCT: A fully automatic artificial intelligence hybrid framework based on convolutional neural network and multi-atlas segmentation for multi-structure cardiac computed tomography angiography image segmentation

Cardiac computed tomography angiography (CTA) is an emerging imaging modality for assessing coronary artery as well as various cardiovascular structures. Recently, deep learning (DL) methods have been successfully applied to many applications of medical image analysis including cardiac CTA structure segmentation. However, DL requires a large amounts of data and high-quality labels for training which can be burdensome to obtain due to its labor-intensive nature. In this study, we aim to develop a fully automatic artificial intelligence (AI) system, named DeepHeartCT, for accurate and rapid cardiac CTA segmentation based on DL. The proposed system was trained using a large clinical dataset with computer-generated labels to segment various cardiovascular structures including left and right ventricles (LV, RV), left and right atria (LA, RA), and LV myocardium (LVM). This new system was trained directly using high-quality computer labels generated from our previously developed multi-atlas based AI system. In addition, a reverse ranking strategy was proposed to assess the segmentation quality in the absence of manual reference labels. This strategy allowed the new framework to assemble optimal computer-generated labels from a large dataset for effective training of a deep convolutional neural network (CNN). A large clinical cardiac CTA studies (n = 1,064) were used to train and validate our framework. The trained model was then tested on another independent dataset with manual labels (n = 60). The Dice score, Hausdorff distance and mean surface distance were used to quantify the segmentation accuracy. The proposed DeepHeartCT framework yields a high median Dice score of 0.90 [interquartile range (IQR), 0.90-0.91], a low median Hausdorff distance of 7 mm (IQR, 4-15 mm) and a low mean surface distance of 0.80 mm (IQR, 0.57-1.29 mm) across all segmented structures. An additional experiment was conducted to evaluate the proposed DL-based AI framework trained with a small vs. large dataset. The results show our framework also performed well when trained on a small optimal training dataset (n = 110) with a significantly reduced training time. These results demonstrated that the proposed DeepHeartCT framework provides accurate and rapid cardiac CTA segmentation that can be readily generalized for handling large-scale medical imaging applications.

A deep-learning approach for direct whole-heart mesh reconstruction

Automated construction of surface geometries of cardiac structures from volumetric medical images is important for a number of clinical applications. While deep-learning-based approaches have demonstrated promising reconstruction precision, these approaches have mostly focused on voxel-wise segmentation followed by surface reconstruction and post-processing techniques. However, such approaches suffer from a number of limitations including disconnected regions or incorrect surface topology due to erroneous segmentation and stair-case artifacts due to limited segmentation resolution. We propose a novel deep-learning-based approach that directly predicts whole heart surface meshes from volumetric CT and MR image data. Our approach leverages a graph convolutional neural network to predict deformation on mesh vertices from a pre-defined mesh template to reconstruct multiple anatomical structures in a 3D image volume. Our method demonstrated promising performance of generating whole heart reconstructions with as good or better accuracy than prior deep-learning-based methods on both CT and MR data. Furthermore, by deforming a template mesh, our method can generate whole heart geometries with better anatomical consistency and produce high-resolution geometries from lower resolution input image data. Our method was also able to produce temporally-consistent surface mesh predictions for heart motion from CT or MR cine sequences, and therefore can potentially be applied for efficiently constructing 4D whole heart dynamics. Our code and pre-trained networks are available at https://github.com/fkong7/MeshDeformNet.

The relationship between systemic inflammation and increased left ventricular mass is partly mediated by noncalcified coronary artery disease burden in psoriasis

Objective

Increased left ventricular (LV) mass is an important precursor to heart failure. Inflammation plays an important role in increasing LV mass. However, the contribution of subclinical coronary artery disease (CAD) to the inflammation-LV mass relationship is unknown. In subjects with psoriasis, a chronic inflammatory skin disease, we evaluated if systemic inflammation assessed by plasma glycoprotein A (GlycA) associated with LV mass measured on coronary CT angiography (CCTA). Additionally, we analyzed whether this relationship was mediated by early CAD assessed as noncalcified coronary burden (NCB).

Methods

We performed an observational longitudinal study of 213 subjects with psoriasis free of known cardiovascular disease, 189 of whom were followed over one year. All participants had GlycA measurements by nuclear magnetic resonance spectroscopy and LV mass and NCB quantified by CCTA.

Results

The cohort had a mean age of 50.3 (±12.9) years and 59% were male. There was moderate psoriasis severity and low cardiovascular risk. LV mass increased by GlycA tertiles [1st tertile:24.6 g/m^2.7(3.8), 2nd tertile:25.5 g/m^2.7(3.8), 3rd tertile:27.7 g/m^2.7(5.5), p<0.001]. Both GlycA (β=0.24, p = 0.001) and NCB (β=0.50, p<0.001) associated with LV mass in models adjusted for age, sex, hypertension, hypertension therapy, lipid therapy, biologic therapy for psoriasis, waist:hip ratio, psoriasis disease duration and severity. In multivariable-adjusted mediation analyses, NCB accounted for 32% of the GlycA-LV mass relationship. Finally, over one year, change in NCB independently associated with change in LV mass (β=0.25, p = 0.002).

Conclusions

Both systemic inflammation and coronary artery NCB were associated with LV mass beyond cardiovascular risk factors in psoriasis. Furthermore, a substantial proportion of the inflammatory-LV mass relationship was mediated by NCB. These findings underscore the possible contribution of early coronary artery disease to the relationship between systemic inflammation and LV mass.

Synthetic breath-hold CT generation from free-breathing CT: a novel deep learning approach to predict cardiac dose reduction in deep-inspiration breath-hold radiotherapy

Deep-inspiration breath-hold radiotherapy (DIBH-RT) to reduce the cardiac dose irradiation is widely used but some patients experience little or no reduction. We constructed and compared two prediction models to evaluate the usefulness of our new synthetic DIBH-CT (sCT) model. Ninety-four left-sided breast cancer patients (training cohort: n = 64, test cohort: n = 30) underwent both free-breathing and DIBH planning. The U-Net-based sCT generation model was developed to create the sCT treatment plan. A linear prediction model was constructed for comparison by selecting anatomical predictors of past literature. The primary prediction outcome is the mean heart dose (MHD) reduction, and the coefficient of determination (R2), root mean square error (RMSE) and mean absolute error (MAE) were calculated. Moreover, we evaluated the heart and lungs contours' similarity and Hounsfield unit (HU) difference between both images. The median MHD reduction was 1.14 Gy in DIBH plans and 1.09 Gy in sCT plans (P = 0.96). The sCT model achieved better performance than the linear model (R2: 0.972 vs 0.450, RMSE: 0.120 vs 0.551, MAE: 0.087 vs 0.412). The organ contours were similar between DIBH-CT and sCT: the median Dice (DSC) and Jaccard similarity coefficients (JSC) were 0.912 and 0.838 for the heart and 0.910 and 0.834 for the lungs. The HU difference in the soft-tissue region was smaller than in the air or bone. In conclusion, our new model can generate the affected CT by breath-holding, resulting in high performance and well-visualized prediction, which may have many potential uses in radiation oncology.

Semi-automatic vessel detection for challenging cases of peripheral arterial disease

OBJECTIVES

Manual or semi-automated segmentation of the lower extremity arterial tree in patients with Peripheral arterial disease (PAD) remains a notoriously difficult and time-consuming task. The complex manifestations of the disease, including discontinuities of the vascular flow channels, the presence of calcified atherosclerotic plaque in close vicinity to adjacent bone, and the presence of metal or other imaging artifacts currently preclude fully automated vessel identification. New machine learning techniques may alleviate this challenge, but require large and reasonably well segmented training data.

METHODS

We propose a novel semi-automatic vessel tracking approach for peripheral arteries to facilitate and accelerate the creation of annotated training data by expert cardiovascular radiologists or technologists, while limiting the number of necessary manual interactions, and reducing processing time. After automatically classifying blood vessels, bones, and other tissue, the relevant vessels are tracked and organized in a tree-like structure for further visualization.

RESULTS

We conducted a pilot (N = 9) and a clinical study (N = 24) in which we assess the accuracy and required time for our approach to achieve sufficient quality for clinical application, with our current clinically established workflow as the standard of reference. Our approach enabled expert physicians to readily identify all clinically relevant lower extremity arteries, even in problematic cases, with an average sensitivity of 92.9%, and an average specificity and overall accuracy of 99.9%.

CONCLUSIONS

Compared to the clinical workflow in our collaborating hospitals (28:40 ± 7:45 [mm:ss]), our approach (17:24 ± 6:44 [mm:ss]) is on average 11:16 [mm:ss] (39%) faster.

Improving multi-atlas cardiac structure segmentation of computed tomography angiography: A performance evaluation based on a heterogeneous dataset

Multi-atlas based segmentation is an effective technique that transforms a representative set of atlas images and labels into a target image for structural segmentation. However, a significant limitation of this approach relates to the fact that the atlas and the target images need to be similar in volume orientation, coverage, or acquisition protocols in order to prevent image misregistration and avoid segmentation fault. In this study, we aim to evaluate the impact of using a heterogeneous Computed Tomography Angiography (CTA) dataset on the performance of a multi-atlas cardiac structure segmentation framework. We propose a generalized technique based upon using the Simple Linear Iterative Clustering (SLIC) supervoxel method to detect a bounding box region enclosing the heart before subsequent cardiac structure segmentation. This technique facilitates our framework to process CTA datasets acquired from distinct imaging protocols and to improve its segmentation accuracy and speed. In a four-way cross comparison based on 60 CTA studies from our institution and 60 CTA datasets from the Multi-Modality Whole Heart Segmentation MICCAI challenge, we show that the proposed framework performs well in segmenting seven different cardiac structures based upon interchangeable atlas and target datasets acquired from different imaging settings. For the overall results, our automated segmentation framework attains a median Dice, mean distance, and Hausdorff distance of 0.88, 1.5 mm, and 9.69 mm over the entire datasets. The average processing time was 1.55 min for both datasets. Furthermore, this study shows that it is feasible to exploit heterogenous datasets from different imaging protocols and institutions for accurate multi-atlas cardiac structure segmentation.