Graph-guided Gaussian Process-based Diagnosis of CVD Severity with Uncertainty Measures.

The severity of coronary artery disease can be assessed invasively using the Fractional Flow Reserve (FFR) index which is a useful diagnostic tool for the clinicians to select the treatment approach. The present work capitalizes a Gaussian process (GP) framework over graphs for the prediction of FFR index using only non-invasive imaging and clinical features. More specifically, taking the per-node one-hop connectivity vector as input, we employed a regression-based task by applying an ensemble of graph-adapted Gaussian process experts, with a data-adaptive fashion via online training. The main novelty of the work lies in the fact that for the first time in a medical field the inference model considers only the similarity condition of the patients, instead of their features. Our results demonstrate the impressive merits of the proposed medical EGP (MedEGP) method, in comparison to the single GP, and Linear Regression (LR) models to predict the FFR index, with well-calibrated uncertainty.Clinical Relevance- This paper establishes an accurate non-invasive approach to predict the FFR for the diagnosis of coronary artery disease.

Machine Learning Models Predict the Need of Amputation and/or Peripheral Artery Revascularization in Hypertensive Patients Within 7-Years Follow-Up.

Lower extremity amputation and requirement of peripheral artery revascularization are common outcomes of undiagnosed peripheral artery disease patients. In the current work, prediction models for the need of amputation or peripheral revascularization focused on hypertensive patients within seven years follow up are employed. We applied machine learning (ML) models using classifiers such as Extreme Gradient Boost (XGBoost), Random Forest (RF) and Adaptive Boost (AdaBoost), that will allow clinicians to identify the patients at risk of these two endpoints using simple clinical data. We used the non-interventional cohort of the getABI study in the primary care setting, selecting 4,191 hypertensive patients out of 6,474 patients with age over 65 years old and followed up for vascular events or death up to 7 years. During this follow up period, 150 patients underwent either amputation or peripheral revascularization or both. Accuracy, Specificity, Sensitivity and Area under the receiver operating characteristic curve (AUC) were estimated for each machine learning model. The results demonstrate Random Forest as the most accurate model for the prediction of the composite endpoint in hypertensive patients within 7 years follow-up, achieving 73.27 % accuracy.Clinical Relevance-This study assists clinicians to better predict and treat these serious outcomes, amputation and peripheral revascularization in hypertensive patients.

Cleverballoon: An integrated approach for developing a drug-coated balloon with everolimus.

Coronary artery disease (CAD) is a chronic disease associated with high mortality and morbidity. Although treatment with drug-eluting stents is the most frequent interventional approach for coronary artery disease, drug-coated balloons (DCBs) constitute an innovative alternative, especially in the presence of certain anatomical conditions in the local coronary vasculature. DCBs allow the fast and homogenous transfer of drugs into the arterial wall, during the balloon inflation. Their use has been established for treating in-stent restenosis caused by stent implantation, while recent clinical trials have shown a satisfactory efficacy in de novo small-vessel disease. Several factors affect DCBs performance including the catheter design, the drug dose and formulation. Cleverballoon focuses on the design and development of an innovative DCB with everolimus. For the realization of the development of this new DCB, an integrated approach, including in- vivo, in-vitro studies and in-silico modelling towards the DCB optimization, is presented.Clinical Relevance-The proposed study introduces the integration of in- vivo, in-vitro and in silico approaches in the design and development process of a new DCB, following the principles of 3R's for the replacement, reduction, and refinement of animal and clinical studies.

TAXINOMISIS: A cloud - based platform for risk profiling and patient specific management of the carotid artery disease.

Carotid Artery Disease is a complex multi-disciplinary medical condition causing strokes and several other disfunctions to individuals. Within this work, a cloud - based platform is proposed for clinicians and medical doctors that provides a comprehensive risk assessment tool for carotid artery disease. It includes three modeling levels: baseline data-driven risk assessment, blood flow simulations and plaque progression modeling. The proposed models, which have been validated through a wide set of studies within the TAXINOMISIS project, are delivered to the end users through an easy-to-use cloud platform. The architecture and the deployment of this platform includes interfaces for handling the electronic patient record, the 3D arterial reconstruction, blood flow simulations and risk assessment reporting. TAXINOMISIS, compared with both similar software approaches and with the current clinical workflow, assists clinicians to treat patients more effectively and more accurately by providing innovative and validated tools.Clinical Relevance - Asymptomatic carotid artery disease is a prevalent condition that affects a significant portion of the population, leading to an increased risk of stroke and other cardiovascular events. Early detection and appropriate treatment of this condition can significantly reduce the risk of adverse outcomes and improve patient outcomes. The development of a software tool to assist clinicians in the assessment and management of asymptomatic patients with carotid artery disease is therefore of great clinical relevance. By providing a comprehensive and reliable assessment of the disease and its risk factors, this tool will enable clinicians to make informed decisions regarding patient management and treatment. The impact of this tool on patient outcomes and the reduction of healthcare costs will be of great importance to both patients and the healthcare system.

An All-in-One Tool for 2D Atherosclerotic Disease Assessment and 3D Coronary Artery Reconstruction.

Diagnosis of coronary artery disease is mainly based on invasive imaging modalities such as X-ray angiography, intravascular ultrasound (IVUS) and optical coherence tomography (OCT). Computed tomography coronary angiography (CTCA) is also used as a non-invasive imaging alternative. In this work, we present a novel and unique tool for 3D coronary artery reconstruction and plaque characterization using the abovementioned imaging modalities or their combination. In particular, image processing and deep learning algorithms were employed and validated for the lumen and adventitia borders and plaque characterization at the IVUS and OCT frames. Strut detection is also achieved from the OCT images. Quantitative analysis of the X-ray angiography enables the 3D reconstruction of the lumen geometry and arterial centerline extraction. The fusion of the generated centerline with the results of the OCT or IVUS analysis enables hybrid coronary artery 3D reconstruction, including the plaques and the stent geometry. CTCA image processing using a 3D level set approach allows the reconstruction of the coronary arterial tree, the calcified and non-calcified plaques as well as the detection of the stent location. The modules of the tool were evaluated for efficiency with over 90% agreement of the 3D models with the manual annotations, while a usability assessment using external evaluators demonstrated high usability resulting in a mean System Usability Scale (SUS) score equal to 0.89, classifying the tool as "excellent".

Machine Learning Models for Cardiovascular Disease Events Prediction.

Cardiovascular diseases (CVDs) are among the most serious disorders leading to high mortality rates worldwide. CVDs can be diagnosed and prevented early by identifying risk biomarkers using statistical and machine learning (ML) models, In this work, we utilize clinical CVD risk factors and biochemical data using machine learning models such as Logistic Regression (LR), Support Vector Machine (SVM), Random Forest (RF), Naïve Bayes (NB), Extreme Grading Boosting (XGB) and Adaptive Boosting (AdaBoost) to predict death caused by CVD within ten years of follow-up. We used the cohort of the Ludwigshafen Risk and Cardiovascular Health (LURIC) study and 2943 patients were included in the analysis (484 annotated as dead due to CVD). We calculated the Accuracy (ACC), Precision, Recall, F1-Score, Specificity (SPE) and area under the receiver operating characteristic curve (AUC) of each model. The findings of the comparative analysis show that Logistic Regression has been proven to be the most reliable algorithm having accuracy 72.20 %. These results will be used in the TIMELY study to estimate the risk score and mortality of CVD in patients with 10-year risk.

Detection of Asymptomatic Carotid Artery Stenosis through Machine Learning.

Carotid artery disease, the pathological condition of carotid arteries, is considered as the most significant cause of cerebral events and stroke. Carotid artery disease is considered as an inflammatory process, which involves the deposition and accumulation of atherosclerotic plaque inside the carotid intima, resulting in the narrowing of the arteries. Carotid artery stenosis (CAS) is either symptomatic or asymptomatic and its presence and location is determined by different imaging modalities, such as the carotid duplex ultrasound, the computed tomography angiography, the magnetic resonance angiography (MRA) and the cerebral angiography. The aim of this study is to present a machine learning model for the diagnosis and identification of individuals of asymptomatic carotid artery stenosis, using as input typical health data. More specifically, the overall model is trained with typical demographics, clinical data, risk factors and medical treatment data and is able to classify the individuals into high risk (Class 1-CAS group) and low risk (Class 0-non CAS group) individuals. In the presented study, we implemented a statistical analysis to check the data quality and the distribution into the two classes. Different feature selection techniques, in combination with classification schemes were applied for the development of our machine learning model. The overall methodology has been trained and tested using 881 cases (443 subjects in low risk class and 438 in high risk class). The highest accuracy 0.82 and an area under curve 0.9 were achieved using the relief feature selection technique and the random forest classification scheme.

Investigation of crimping effects on the stent deployment through in silico modeling.

Atherosclerosis is one of the most mortal diseases that affects the arterial vessels, due to accumulation of plaque, altering the hemodynamic environment of the artery by preventing the sufficient delivery of blood to other organs. Stents are expandable tubular wires, used as a treatment option. In silico studies have been extensively exploited towards examining the performance of such devices by employing Finite Element Modeling. This study models the crimping stage during stent implantation to examine the effect of inclusion of pre-stress state of the stent. The results show that modeling of the crimping stress state of the stent prior to the deployment results in under-expansion of the stent, due to the indirect inclusion of strain-induced hardening effects. As a result, it is evident that the compressive stent stress configuration is important to be considered in the computational modeling approaches of stent deployment.

A proof-of-concept study for the simulation of blood flow in a post arterial segment for different blood rheology models.

Cardiovascular disease (CVD) and especially atherosclerosis are chronic inflammatory diseases which cause the atherosclerotic plaque growth in the arterial vessels and the blood flow reduction. Stents have revolutionized the treatment of this disease to a great extent by restoring the blood flow in the vessel. The present study investigates the performance of the blood flow after stent implantation in patient-specific coronary artery and demonstrates the effect of using Newtonian vs. non-Newtonian blood fluid models in the distribution of endothelial shear stress. In particular, the Navier-Stokes and continuity equations were employed, and three non-Newtonian fluid models were investigated (Carreau, Carreau-Yasuda and the Casson model). Computational finite elements models were used for the simulation of blood flow. The comparison of the results demonstrates that the Newtonian fluid model underestimates the calculation of Endothelial Shear Stress, while the three non-Newtonian fluids present similar distribution of shear stress. Keywords: Blood flow dynamics, stented artery, non-Newtonian fluid. Clinical Relevance- This work demonstrates that when blood flow modeling is performed at stented arteries and predictive models are developed, the non-Newtonian nature of blood must be considered.

Segmentation of left atrium using CT images and a deep learning model.

The left atrium (LA) is one of the cardiac cavities with the most complex anatomical structures. Its role in the clinical diagnosis and patient's management is critical, as it is responsible for the atrial fibrillation, a condition that promotes the thrombogenesis inside the left atrial appendage. The development of an automated approach for LA segmentation is a demanding task mainly due to its anatomical complexity and the variation of its shape among patients. In this study, we focus to develop an unbiased pipeline capable to segment the atrial cavity from CT images. For evaluation purposes state-of-the-art metrics were used to assess the segmentation results. Particularly, the results indicated the mean values of the dice score 80%, the hausdorff distance 11.78mm, the average surface distance 2.24mm and the rand error index 0.2.

A Machine Learning Model for the Identification of High risk Carotid Atherosclerotic Plaques.

Carotid artery disease is an inflammatory condition involving the deposition and accumulation of lipid species and leucocytes from blood into the arterial wall, which causes the narrowing of the carotid arteries on either side of the neck. Different imaging modalities can by implemented to determine the presence and the location of carotid artery stenosis, such as carotid ultrasound, computed tomography angiography (CTA), magnetic resonance angiography (MRA), or cerebral angiography. However, except of the presence and the degree of stenosis of the carotid arteries, the vulnerability of the carotid atherosclerotic plaques constitutes a significant factor for the progression of the disease and the presence of disease symptoms. In this study, our aim is to develop and present a machine learning model for the identification of high risk plaques using non imaging based features and non-invasive imaging based features. Firstly, we implemented statistical analysis to identify the most statistical significant features according to the defined output, and subsequently, we implemented different feature selection techniques and classification schemes for the development of our machine learning model. The overall methodology has been trained and tested using 208 cases of 107 cases of low risk plaques and 101 cases of high risk plaques. The highest accuracy of 0.76 was achieved using the relief feature selection technique and the support vector machine classification scheme. The innovative aspect of the proposed machine learning model is both the different categories of the utilized input features and the definition of the problem to be solved.

Computational modeling of atherosclerotic plaque progression in carotid lesions with moderate degree of stenosis.

Carotid atherosclerotic plaque growth leads to the progressive luminal stenosis of the vessel, which may erode or rupture causing thromboembolism and cerebral infarction, manifested as stroke. Carotid atherosclerosis is considered the major cause of ischemic stroke in Europe and thus new imaging-based computational tools that can improve risk stratification and management of carotid artery disease patients are needed. In this work, we present a new computational approach for modeling atherosclerotic plaque progression in real patient-carotid lesions, with moderate to severe degree of stenosis (>50%). The model incorporates for the first time, the baseline 3D geometry of the plaque tissue components (e.g. Lipid Core) identified by MR imaging, in which the major biological processes of atherosclerosis are simulated in time. The simulated plaque tissue production results in the inward remodeling of the vessel wall promoting luminal stenosis which in turn predicts the region of the actual stenosis progression observed at the follow-up visit. The model aims to support clinical decision making, by identifying regions prone to plaque formation, predict carotid stenosis and plaque burden progression, and provide advice on the optimal time for patient follow-up screening.

A proof-of-concept study for the prediction of the de-novo atherosclerotic plaque development using finite elements.

The type of the atherosclerotic plaque has significant clinical meaning since plaque vulnerability depends on its type. In this work, we present a computational approach which predicts the development of new plaques in coronary arteries. More specifically, we employ a multi-level model which simulates the blood fluid dynamics, the lipoprotein transport and their accumulation in the arterial wall and the triggering of inflammation using convection-diffusion-reaction equations and in the final level, we estimate the plaque volume which causes the arterial wall thickening. The novelty of this work relies on the conceptual approach that using the information from 94 patients with computed tomography coronary angiography (CTCA) imaging at two time points we identify the correlation of the computational results with the real plaque components detected in CTCA. In the next step, we use these correlations to generate two types of de-novo plaques: calcified and non-calcified. Evaluation of the model's performance is achieved using eleven patients, who present de-novo plaques at the follow-up imaging. The results demonstrate that the computationally generated plaques are associated significantly with the real plaques indicating that the proposed approach could be used for the prediction of specific plaque type formation.

Investigation of Drug Eluting Stents performance in human atherosclerotic artery through in silico modeling.

Atherosclerosis is a chronic inflammatory disease associated with heart attack and stroke. It causes the growth of atherosclerotic plaques inside the arterial vessels, which in turn results to the reduction of the blood flow to the different organs. Drug-Eluting Stents (DES) are mesh-like wires, carrying pharmaceutical coating, designed to dilate and support the arterial vessel, restore blood flow and through the controlled local drug delivery inhibit neo-intimal thickening. In silico modeling is an efficient method of accurately predicting and assessing the performance of the stenting procedure. The present in silico study investigates the performance of two different stents (Bare Metal Stent, Drug-Eluting Stent) in a patient-specific coronary artery and assesses the effect of stent coating, considering that the same procedural approach is followed by the interventional cardiologist. The results demonstrate that even if small differences are obtained in the two models, the incorporation of the stent coatings (in DES) does not significantly affect the outcomes of the stent deployment, the stresses and strains in the scaffold and the arterial tissue. Nevertheless, it is suggested that regarding the DES expansion, higher pressure should be applied at the inner surface of the stent.

Error Propagation in the Simulation of Atherosclerotic Plaque Growth and the Prediction of Atherosclerotic Disease Progression.

Assessments of coronary artery disease can be achieved using non-invasive computed tomography coronary angiography (CTCA). CTCA can be further used for the 3D reconstruction of the coronary arteries and the development of computational models. However, image acquisition and arterial reconstruction introduce an error which can be propagated, affecting the computational results and the accuracy of diagnostic and prognostic models. In this work, we investigate the effect of an imaging error, propagated to a diagnostic index calculated using computational modelling of blood flow and then to prognostic models based on plaque growth modelling or binary logistic predictive modelling. The analysis was performed utilizing data from 20 patients collected at two time points with interscan period of six years. The collected data includes clinical and risk factors, biological and biohumoral data, and CTCA imaging. The results demonstrated that the error propagated and may have significantly affected some of the final outcomes. The calculated propagated error seemed to be minor for shear stress, but was major for some variables of the plaque growth model. In parallel, in the current analysis SmartFFR was not considerably affected, with the limitation of only one case located into the gray zone.

SmartFFR, a New Functional Index of Coronary Stenosis: Comparison With Invasive FFR Data.

Aims: In this study, we evaluate the efficacy of SmartFFR, a new functional index of coronary stenosis severity compared with gold standard invasive measurement of fractional flow reserve (FFR). We also assess the influence of the type of simulation employed on smartFFR (i.e. Fluid Structure Interaction vs. rigid wall assumption). Methods and Results: In a dataset of 167 patients undergoing either computed tomography coronary angiography (CTCA) and invasive coronary angiography or only invasive coronary angiography (ICA), as well as invasive FFR measurement, SmartFFR was computed after the 3D reconstruction of the vessels of interest and the subsequent blood flow simulations. 202 vessels were analyzed with a mean total computational time of seven minutes. SmartFFR was used to process all models reconstructed by either method. The mean FFR value of the examined dataset was 0.846 ± 0.089 with 95% CI for the mean of 0.833-0.858, whereas the mean SmartFFR value was 0.853 ± 0.095 with 95% CI for the mean of 0.84-0.866. SmartFFR was significantly correlated with invasive FFR values (RCCTA = 0.86, p CCTA < 0.0001, RICA = 0.84, p ICA < 0.0001, R overall = 0.833, p overall < 0.0001), showing good agreement as depicted by the Bland-Altman method of analysis. The optimal SmartFFR threshold to diagnose ischemia was ≤0.83 for the overall dataset, ≤0.83 for the CTCA-derived dataset and ≤0.81 for the ICA-derived dataset, as defined by a ROC analysis (AUCoverall = 0.956, p < 0.001, AUCICA = 0.975, p < 0.001, AUCCCTA = 0.952, p < 0.001). Conclusion: SmartFFR is a fast and accurate on-site index of hemodynamic significance of coronary stenosis both at single coronary segment and at two or more branches level simultaneously, which can be applied to all CTCA or ICA sequences of acceptable quality.

Cardiac CT perfusion and FFRCTA: pathophysiological features in ischemic heart disease.

Cardiac computed tomography (CCT) has rapidly evolved, becoming a powerful integrated tool for the evaluation of coronary artery disease (CAD), and being superior to other noninvasive methods due to its high accuracy and ability to simultaneously assess both lumen stenosis and atherosclerotic plaque burden. Furthermore, CCT is regarded as an effective gatekeeper for coronary angiography, and carries independent important prognostic information. In the last decade, the introduction of new functional CCT applications, namely CCT perfusion (CCTP) imaging and CT-derived fractional flow reserve (FFRCTA), has opened the door for accurate assessment of the haemodynamic significance of stenoses. These new CCT technologies, thus, share the unique advantage of assessing both myocardial ischemia and patient-specific coronary artery anatomy, providing an integrated anatomical/functional analysis. In the present review, starting from the pathophysiology of myocardial ischemia, we evaluate the existing evidence for functional CCT imaging and its value in relation to alternative, well-established, non-invasive imaging modalities and invasive indices of ischemia (currently the gold-standard). The knowledge of clinical applications, benefits, and limitations of these new CCT technologies will allow efficient and optimal use in clinical practice in the near future.

Validation study of a novel method for the 3D reconstruction of coronary bifurcations.

Quantitative Coronary Angiography (QCA) is an important tool in the study of coronary artery disease. Validation of this technique is crucial for their ongoing development and refinement although it is difficult due to several factors such as potential sources of error. The present work aims to a further validation of a new semi-automated method for three-dimensional (3D) reconstruction of coronary bifurcations arteries based on X-Ray Coronary Angiographies (CA). In a dataset of 40 patients (79 angiographic views), we used the aforementioned method to reconstruct them in 3D space. The validation was based on the comparison of these 3D models with the true silhouette of 2D models annotated by an expert using specific metrics. The obtained results indicate a good accuracy for the most parameters (≥ 90 %). Comparison with similar works shows that our new method is a promising tool for the 3D reconstruction of coronary bifurcations and for application in everyday clinical use.

Wall shear stress alterations at left atrium and left atrial appendage employing abnormal blood velocity profiles.

In this study, we developed and analyzed different patient-specific 3D anatomical models of the left atrium including left atrial Appendage, in order to investigate the local hemodynamics. Particularly, we focused on the left atrial appendage and its impact on thrombus formation due to wall shear stress alterations. A 3D semi-automated reconstruction approach was carried out to segment and reconstruct the left atrium from CT scans. Six different patients were studied applying their patient-specific clinical data. Three different velocity profiles simulated for each patient case, representing one normal and two abnormal conditions. Simulations varied significantly according to different appendage morphologies. Our scope is to describe the hemodynamic behavior at the left atrium and the left atrial appendage according to different blood velocities based on their anatomic variety (chicken wing 0.14 m/s, windsock 0.10, cactus 0.08, and cauliflower 0.04). Wall shear stress results were demonstrated and correlated with the velocities and the thrombus formation inside the appendage cavity.

Prediction of atherosclerotic disease progression combining computational modelling with machine learning.

Non-invasive serial computed tomography coronary angiography (CTCA) was acquired from 32 patients and 3D reconstruction of 58 coronary arteries was achieved. The arterial geometries were utilized for blood flow and LDL transport modelling. Navier-Stokes and convection-diffusion equations were employed for simulation of blood flow and LDL transport, respectively. Disease progression was assessed comparing the follow-up and baseline arterial models after co-registration using side branches as anatomical landmarks. A machine learning model for predicting disease progression was built using the Gradient Boosted Trees (GBT) algorithm. The Accuracy, Sensitivity, Specificity and AUC of the developed methodology for predicting lumen area decrease equal was 0.68, 0.56, 0.34 and 0.59, respectively. The best results were found for the prediction of plaque area increase by 20%, with 0.73, 0.67, 0.86, and 0.76 accuracy, sensitivity, specificity andAUC, respectively. This approach outperforms significantly the predictive capability of models based on binary logistic regression.