Cardiovascular disease (CVD) remains a leading global health burden, with carotid intima-media thickness (cIMT) recognized as a sensitive, non-invasive biomarker for early atherosclerosis and future cardiovascular risk. Although ultrasound imaging is the standard method for measuring cIMT, its accessibility may be limited in large-scale populations with intensive screening demands or in low-resource settings, posing a significant challenge for stroke survivors who have a high need for CVD assessment. While existing cIMT prediction models based solely on tabular data have been proposed to bypass the need for strong reliance on image-derived features, they typically frame the task as a binary classification problem, indicating only the presence or absence of vascular risk, thereby failing to effectively capture its actual severity. In contrast, this work proposes a prognostic learning model to effectively estimate cIMT, enabling precise quantification of atherosclerosis severity without relying on imaging data. By constructing a patient similarity graph relying on demographic and clinical-derived features to (i) bypass the need for revealing the actual clinical measurements, promoting privacy, and (ii) to explicitly account for patient's interdependencies, this work introduces a graph-guided self-supervised learning (Self-SL) framework to learn informative representations for the cIMT prediction task. These learned representations encode key local and global graph information that can readily assist the downstream task requiring only a minimal amount of labeled data. Applied to the UK Biobank cohort, the model outperforms conventional learning models, achieving up to 93.22% average MSE reduction, underscoring graph similarity strength in capturing latent clinical patterns.

Cardiac rehabilitation (CR) programs are vital for people recovering from cardiac surgeries or events. However, the effectiveness of CR programs varies and some patients may not adhere to them, which might result in less favourable outcomes. Machine Learning (ML) models could help predict the effectiveness and adherence of CR programs. This study proposes two such models: a) the CR program effectiveness prediction model and b) the CR program adherence prediction model. The models were trained on data from retrospective cohort study with 1448 participants collected at the Cardiac Rehabilitation Unit of the Hospital Clinico de Santiago de Compostela in Galicia, Spain (SERGAS). Data cleaning, normalization, imputation, statistical analysis, feature selection and repeated stratified k-fold cross-validation (CV) were applied on the ML pipeline, which tested and evaluated on baseline demographic, clinical, exercise tests and behavioral features. The performance of ML models was assessed by mean Area Under operating characteristic Curve (AUC), specificity, sensitivity, and balanced accuracy with 95% confidence interval (CI). The results show that Random Forest (RF) outperformed other evaluated classifiers for the CR program effectiveness model, with the highest AUC value of 0.789 (0.775, 0.802), while the best classifier for the CR adherence model was the Logistic Regression (LR) classifier, with an AUC value of 0.757 (0.749, 0.764). SHAP plots were also used to investigate the relationships among the variables used in the analysis. Finally, a two-dimensional scoring system was developed to jointly assess predicted adherence and effectiveness, enabling personalized visualization of patient response to CR.

Heart failure (HF) is a complex syndrome that is affected by many factors and causes. It is crucial to early recognize the disease subtypes and the unidentified clinical pathways that give rise to it. Machine learning (ML) is the tool that assist to deal with these challenges and improve the prediction of HF. In this work, the HF risk prediction was implemented by employed ML classifiers, such as Random Forest (RF), Extreme Grading Boosting (XGBoost) and Light Gradient-Boosting Machine (LGBM). We utilized the data from the German epidemiological trial on ankle brachial index - getABI cohort, which includes 6,454 patients. The performance of classifiers was estimated by Accuracy (ACC), Sensitivity, Specificity and the area under the receiver operating characteristic curve (AUC) in mean values for each ML classifier. The results were also interpreted using the Explainable artificial intelligence (XAI) approach, the Shapley Additive exPlanations (SHAP) values. Our work reveals that LGBM classifier predict the HF risk within 5 years follow-up in general population with 68 % accuracy. Moreover, the N-terminal pro-B-type natriuretic peptide (NT-proBNP) was identified as the most important feature for HF risk prediction.

Fatal myocardial infarction (MI) is one of the most common types of cardiovascular diseases that often presents in the emergency department. The prediction of death caused by myocardial infarction within 10-years follow-up is addressed in this study, using comorbidities, daily habits, clinical and laboratory data in binary and continuous data respectively. The used data are included in a cohort of the Ludwigshafen Risk and Cardiovascular Health (LURIC) study. The target feature, namely death caused by MI, contained 106 deceased patients and 2,321 alive patients in the final used dataset. The analysis was based on machine learning models (ML), such as support vector machine (SVM), light gradient-boosting machine (LGBM), Random Forest (RF), Decision Tree (DT). Their performance was estimated by Area Under Receiver Operating Characteristic Curve (AUC), Sensitivity, Specificity, Precision and Accuracy. Results show that LGBM was the most suitable of the aforementioned models to predict death caused by myocardial infarction within 10-years follow-up, achieving area under the curve (AUC) value equal to 77.03 %, accuracy 69.42 %, sensitivity 69.75 %, specificity 69.40% and precision 53.76 %. In addition, explainable artificial intelligence (xAI) was utilized and especially SHapley Additive exPlanations (SHAP) was the selected method. SHAP was utilized in order to shed light on the results, applying the LGBM model as the best predictive model. The provided SHAP plots contribute to the interpretation of how each independent feature aid to the final prediction.

Cardiovascular disease (CVD) poses a significant global health challenge, and accurate inference methods are vital for early detection and intervention. However, the quality of prediction relies heavily on the availability of labeled data, which are often limited in medical applications. To cope with the challenge of limited labeled data, we are the first to propose an active learning (AL) approach that leverages a weighted ensemble of Gaussian processes to effectively infer CVD by strategically selecting the few most informative data points to label. Through experiments conducted on the SMARTool dataset, we demonstrate the effectiveness of the advocated approach, achieving superior performance in CVD inference compared to baseline methods. Our findings highlight the potential impact of the proposed AL framework in CVD diagnosis and treatment clinical cases, particularly in scenarios where labeled data are scarce, due to data confidentiality concerns or high sampling costs.

Coronary artery disease (CAD) is one of the leading causes of mortality worldwide. Drug-eluting stents (DES) are nowadays widely used so as to treat the occluded arteries, restore blood flow and through the diffusion of the drug achieve better clinical outcomes compared to Bare Metal Stents (BMSs), in terms of reduced numbers of cardiac death, myocardial infarction and vessel revascularization. BioCoStent targets the design and development of an innovative DES with retinoic acid. In this study the overall concept for realizing this new DES development, including the characterization of the biomaterials, the performance of in vivo and in vitro studies and the optimisation through in silico modelling, is presented.

In this work, we present a novel and accurate methodology for the segmentation of optical coherence tomography imaging (OCT) and detection of lumen and outer wall, plaque characterization and stent struts in stented arteries. In particular, the methodology starts with pre-processing and detection of the catheter artefact. Struts detection is based on the identification of the size of the shadow behind the struts. Our methodology can be applied to metal stents as well as to polymeric and bioresorbable vascular scaffold (BVS) stents. Lumen segmentation is based on Fuzzy clustering and Fast marching on the gradient image to find the shortest path. The outer wall is segmented using a methodology, which combines K-means and 3-dimensional (3D) surface fitting on the detected edges. K-means with 3 clusters is performed at the final step on the ROI between the lumen and outer border of adventitia to characterize the plaque type. The validation is achieved by comparing the algorithm's results with manual annotations provided by experts. The results demonstrate that our methodology is accurate in lumen (R=0.99) and outer wall segmentation (R=0.77) and struts detection (R=0.82). The average Hausdorff distance and the Dice Similarity for lumen segmentation is 0.097 mm and 0.96, respectively.

Atherosclerosis is the one of the major causes of mortality worldwide, urging the need for its treatment. This study is aiming to investigate the role of diabetes in the atherosclerotic plaque growth mechanisms through the utilization of a multi-level numerical model. To accomplish this, we developed a proof-of-concept mathematical model of the diabetes effect to plaque growth, that has been coupled to a stateof-the-art multi-level numerical model of plaque growth. Diabetes main effect is the increase of the average blood glucose concentration, which causes the decrease of the endothelial nitric oxide production rate by affecting several biologic pathways. Nitric oxide is a signaling molecule that regulates the endothelial flow rates, and any abnormal alteration leads to endothelial dysfunction, the major culprit of atherosclerosis. The derived model considers the modeling of blood flow in lumen and of species transport and reactions in the arterial wall. The considered factors include: (i) LDL, (ii) HDL, (iii) oxidized LDL, (iv) monocytes, (v) macrophages, (vi) cytokines, (vii) smooth muscle cells (contractile & synthetic), and (viii) collagen. The model is validated using 10 patients' reconstructed arterial data in two time-points. More specifically, baseline geometries are used as an input to our model, while follow-up geometries are used as benchmark for our model's output. The results presented a high coefficient of determination between the simulated with diabetes effect and the real follow-up geometries of 0.634.

Coronary stents are expandable scaffolds that are used to widen occluded diseased arteries and restore blood flow. Because of the strain they are exposed to and forces they must resist as well as the importance of surface interactions, material properties are dominant. Indeed, a common differentiating factors amongst commercially available stents is their material. Several performance requirements relate to stent materials including radial strength for adequate arterial support post-deployment. This study investigated the effect of the stent material in three finite element models using different stents made of: (i) Cobalt-Chromium (CoCr), (ii) Stainless Steel (SS316L), and (iii) Platinum Chromium (PtCr). Deployment was investigated in a patient specific arterial geometry, created based on a fusion of angiographic data and intravascular ultrasound images. In silico results show that: (i) the maximum von Mises stress occurs for the CoCr, however the curved areas of the stent links present higher stresses compared to the straight stent segments for all stents, (ii) more areas of high inner arterial stress exist in the case of the CoCr stent deployment, (iii) there is no significant difference in the percentage of arterial stress volume distribution among all models.

Coronary Computed Tomography Angiography (CCTA) has gained substantial ground in everyday clinical practice due to its non-invasive nature. In this work we present a noninvasive method to assess the hemodynamic significance of coronary stenoses using only CCTA images. Two female patients were subjected to Invasive Coronary Angiography, Virtual Histology IVUS and CCTA. The same arterial segment was reconstructed in 3D using the proposed method as well as two already validated 3D reconstruction methods using the aforementioned invasive techniques. The lumen diameter reduction (%) and the minimum lumen diameter (mm) were calculated for all cases and a relative error <;5% was observed between all three techniques.

Although micro-anastomosis is the most commonly performed procedure for reconnecting two blood vessels through sutures, thrombus formation and subsequently anastomotic failure remains one of the most serious clinical complications. An important stimulus to thrombus formation is the altered hemodynamics with abnormal Wall Shear Stress (WSS) distribution on endothelial cells generated by the presence of sutures. Computational simulation is a valid tool to examine the local hemodynamics of micro-anastomosed vessels, allowing for the calculation of the WSS, a factor that could otherwise not directly possible to be measured in vivo. The aim of this study is to perform Fluid-Structure Interaction (FSI) analysis of micro-anastomosis in order to examine the effects of the wall compliance on the hemodynamic quantities.

The aim of this study is to investigate the effect of mallear and incudal ligaments to the tympanic membrane and the stapes footplate displacement in a finite element model of the middle ear. Three cases were simulated: one without the ligaments, one including the posterior incudal and the anterior mallear ligaments and one including in addition the superior mallear and incudal ligaments. A maximum stapes footplate displacement 0.023 μm was observed at a frequency 1024 Hz by exciting the tympanic membrane at a sinusoidal sound pressure level (SPL) of 90 dB. The computational results were validated with experimental measurements from the literature. Concluding our results show that the superior ligaments are most beneficial for an accurate representation of the middle ear frequency response. Excellent agreement is observed between our results and human temporal bone experimental data and other finite element studies.

An unsupervised methodology for the detection of Epileptic seizures in EEG recordings is proposed. The time-frequency content of the EEG signals is extracted using the Short Time Fourier Transform. The analysis focuses on the EEG energy distribution among the well-established delta, theta and alpha rhythms (2-13 Hz), as energy variations in these frequency bands are widely associated with seizure activity. Relying on seizure rhythmicity, the classification is performed by isolating the segments where each rhythm is more clearly and dominantly expressed over the others. For the first time, an unsupervised methodology is evaluated using more than 978 hours of EEG recordings from a public database. The results show that the proposed methodology achieves high seizure detection sensitivity with significantly reduced human intervention.

Progression of atherosclerotic process constitutes a serious and quite common condition due to accumulation of fatty materials in the arterial wall, consequently posing serious cardiovascular complications. In this paper, we assemble and analyze a multitude of heterogeneous data in order to model the progression of atherosclerosis (ATS) in coronary vessels. The patient's medical record, biochemical analytes, monocyte information, adhesion molecules, and therapy-related data comprise the input for the subsequent analysis. As indicator of coronary lesion progression, two consecutive coronary computed tomography angiographies have been evaluated in the same patient. To this end, a set of 39 patients is studied using a twofold approach, namely, baseline analysis and temporal analysis. The former approach employs baseline information in order to predict the future state of the patient (in terms of progression of ATS). The latter is based on an approach encompassing dynamic Bayesian networks whereby snapshots of the patient's status over the follow-up are analyzed in order to model the evolvement of ATS, taking into account the temporal dimension of the disease. The quantitative assessment of our work has resulted in 93.3% accuracy for the case of baseline analysis, and 83% overall accuracy for the temporal analysis, in terms of modeling and predicting the evolvement of ATS. It should be noted that the application of the SMOTE algorithm for handling class imbalance and the subsequent evaluation procedure might have introduced an overestimation of the performance metrics, due to the employment of synthesized instances. The most prominent features found to play a substantial role in the progression of the disease are: diabetes, cholesterol and cholesterol/HDL. Among novel markers, the CD11b marker of leukocyte integrin complex is associated with coronary plaque progression.

Stents are medical devices used in cardiovascular intervention for unblocking the diseased arteries and restoring blood flow. During stent implantation the deformation of the arterial wall as well as the resulted stresses caused in the arterial morphology are studied. In this paper we study the effect of the composition of the atherosclerotic plaque during the stent deployment procedure, using Finite Element modeling. The stenting procedure is simulated for two different cases; in the first the presence of the plaque is ignored whereas in the second a three dimensional (3D) stiff calcified plaque is located in the stenotic area of the artery. Results indicate that in the second case the von Mises stresses in the arterial wall are higher than the stresses occurred in the first case. In addition, the distribution of the arterial von Mises stress depends on the plaque composition.

In this study, we present an innovative model for plaque growth utilizing a 3-Dimensional (3D) left coronary arterial tree reconstructed from computed tomographic (CT) data. The proposed model takes into consideration not only the effect of the local hemodynamic factors but also major biological processes such as the low density lipoprotein (LDL) and high density lipoprotein (HDL) transport, the macrophages recruitment and the foam cells formation. The endothelial membrane is considered semi-permeable and endothelial shear stress dependent, while its permeability is modeled using the Kedem-Katscalsky equations. Patient specific biological data are used for the accurate modeling of plaque formation process. The finite element method (FEM) is employed for the solution of the system of partial differential equations. The results of the simulation are compared to the plaque progression in a follow-up CT examination performed three years after the initial investigation. The results show that the proposed model can be used to predict regions prone for plaque development of progression.

Computational fluid dynamics methods based on in vivo 3-D vessel reconstructions have recently been identified the influence of wall shear stress on endothelial cells as well as on vascular smooth muscle cells, resulting in different events such as flow mediated vasodilatation, atherosclerosis, and vascular remodeling. Development of image-based modeling technologies for simulating patient-specific local blood flows is introducing a novel approach to risk prediction for coronary plaque growth and progression. In this study, we developed 3-D model of plaque formation and progression that was tested in a set of patients who underwent coronary computed tomography angiography (CTA) for anginal symptoms. The 3-D blood flow is described by the Navier-Stokes equations, together with the continuity equation. Mass transfer within the blood lumen and through the arterial wall is coupled with the blood flow and is modeled by a convection-diffusion equation. The low density lipoprotein (LDL) transports in lumen of the vessel and through the vessel tissue (which has a mass consumption term) are coupled by Kedem-Katchalsky equations. The inflammatory process is modeled using three additional reaction-diffusion partial differential equations. A full 3-D model was created. It includes blood flow and LDL concentration, as well as plaque formation and progression. Furthermore, features potentially affecting plaque growth, such as patient risk score, circulating biomarkers, localization and composition of the initial plaque, and coronary vasodilating capability were also investigated. The proof of concept of the model effectiveness was assessed by repetition of CTA, six months after the baseline evaluation. Besides the low values of local shear stress, plaque characteristics, risk profile, pattern of circulating adhesion molecules, and reduced coronary flow reserve at baseline appeared to affect plaque progression toward flow-limiting lesions at follow-up evaluation. Although preliminary, our multidisciplinary approach to a “personalized” prediction of coronary plaque progression suggests that incorporation in atherosclerotic models of systemic and local hemodynamic features may better predict evolution of plaques in coronary artery disease stable patients.