Search Results (121)

Background: Mild Cognitive Impairment (MCI) represents an intermediate stage between normal cognitive aging and Alzheimer’s Disease (AD). Early and accurate identification of MCI is crucial for implementing interventions that may delay or prevent further cognitive decline. This study aims to develop a machine learning-based model for differentiating between Cognitively Normal (CN) individuals and MCI patients using data from the Alzheimer’s Disease Neuroimaging Initiative (ADNI). Methods: An ensemble classification approach was designed by integrating Extra Trees, Random Forest, and Light Gradient Boosting Machine (LightGBM) algorithms. Feature selection emphasized clinically relevant biomarkers, including Amyloid-β 42, phosphorylated tau, diastolic blood pressure, age, and gender. The dataset was split into training and held-out test sets. A probability thresholding strategy was employed to flag uncertain predictions for potential deferral, enhancing model reliability in borderline cases. Results: The final ensemble model achieved an accuracy of 83.2%, a recall of 80.2%, and a precision of 86.3% on the independent test set. The probability thresholding mechanism flagged 23.3% of cases as uncertain, allowing the system to abstain from low-confidence predictions. This strategy improved clinical interpretability and minimized the risk of misclassification in ambiguous cases. Conclusions: The proposed AI-driven ensemble model demonstrates strong performance in classifying MCI versus CN individuals using multimodal ADNI data. Incorporating a deferral mechanism through uncertainty estimation further enhances the model’s clinical utility. These findings support the integration of machine learning tools into early screening workflows for cognitive impairment. Full article

Background: Chronic kidney disease (CKD) is a prevalent complication among individuals with type 2 diabetes (T2D), posing significant diagnostic challenges in resource-limited settings due to infrequent testing and missed hospital visits. This study aimed to develop a simple, effective ML model to identify T2D patients at high risk for reduced kidney function. Methods: We retrospectively analyzed data from 3471 T2D patients collected over a ten-year period at a university hospital in Bangkok, Thailand. Two models were developed using readily available clinical features: one including hemoglobin A1c (HbA1c) levels (the “with-HbA1c” model) and one excluding HbA1c levels (the “non–HbA1c” model). Three tree-based ML algorithms—decision tree, random forest, and extreme gradient boosting (XGBoost) algorithms—were employed. The outcome label was CKD, defined as an estimated Glomerular Filtration Rate (eGFR) < 60 mL/min/1.73 m² that persisted for more than 90 days. The model performance was evaluated using the AUROC. The feature importance was assessed using Shapley additive explanations (SHAP). Results: The XGBoost algorithm demonstrated a strong predictive performance. The “with-HbA1c” model achieved an AUROC of 0.824, while the “non–HbA1c” model attained a comparable AUROC of 0.819. Both models were well-calibrated. SHAP analysis identified age, HbA1c, and systolic blood pressure as the most influential predictors. Conclusions: Our simplified, interpretable ML models can effectively stratify the risk of reduced kidney function in patients with T2D using minimal, routine data. These models represent a promising step toward integration into clinical practice, such as through EHR-based alerts or patient-facing mobile applications, to improve early CKD detection, particularly in resource-limited settings. Full article

This study presents an innovative approach to quantifying uncertainty in Herschel–Bulkley (H-B) fluid flow through rectangular ducts, analyzing four scenarios: uncertain apparent viscosity (Case I), uncertain pressure gradient (Case II), uncertain boundary conditions (Case III) and uncertain apparent viscosity and pressure gradient (Case IV). Using the stochastic finite difference with homogeneous chaos (SFDHC) method, we produce probability density functions (PDFs) of fluid velocity with exceptional computational efficiency (243 times faster), matching the accuracy of Monte Carlo simulation (MCS). Key statistics and maximum velocity PDFs are tabulated and visualized for each case. Mean velocity shows minimal variation in Cases I, III, and IV, but maximum velocity fluctuates significantly in Case I (63.95–187.45% of mean), Case II (50.15–156.68%), and Case IV (63.70–185.53% of mean), vital for duct design and analysis. Examining the effects of different parameters, the SFDHC method’s rapid convergence reveals the fluid behavior index as the primary driver of maximum stochastic velocity, followed by aspect ratio and yield stress. These findings enhance applications in drilling fluid management, biomedical modeling (e.g., blood flow in vascular networks), and industrial processes involving non-Newtonian fluids, such as paints and slurries, providing a robust tool for advancing understanding and managing uncertainty in complex fluid dynamics. Full article

Background/Objectives: Aortic valve stenosis (AS) is a prevalent cardiovascular condition among elderly patients frequently treated with Transcatheter Aortic Valve Implantation (TAVI). Traditional hemodynamic monitoring during TAVI relies on invasive methods. The ClearSight^® Finger Cuff system offers a non-invasive alternative for continuous hemodynamic monitoring. To compare the reliability and feasibility of non-invasive hemodynamic monitoring with traditional invasive hemodynamic monitoring during TAVI procedures. Methods: In this prospective observational study, patients undergoing elective TAVI were recruited from two tertiary hospitals between March and August 2023. Invasive hemodynamic measurements were obtained using arterial and pigtail catheters, with a subset undergoing right heart catheterization. Non-invasive measurements were captured using the ClearSight^® system. Data on baseline characteristics, procedural details, and 30-day follow-up outcomes were collected. Results: The study cohort comprised 50 patients (median age 82 years (IQR 78.0, 85.8), 50% female). Non-invasive measurements of cardiac output (CO), cardiac index (CI), and stroke volume (SV) were consistently lower than invasive measurements (CO: 4.1 vs. 4.8 L/min, p = 0.03; CI: 2.2 vs. 2.7 L/min/m², p = 0.01, SV: 66 vs. 77 mL, p = 0.25). Non-invasive blood pressure readings were lower than invasive radial and aortic measurements before and after TAVI. Correlation of non- and invasive measurements was low but similar before and after TAVI (Mean percentage error of 52%). Conclusions: The ClearSight^® system provided lower absolute values for all evaluated hemodynamic parameters as well as low correlation compared to traditional methods pre- as well as post-interventional. Full article

Background and Objectives: After transcatheter aortic valve implantation (TAVI), transvalvular gradients increase immediately following the procedure up to 24 h afterward. While factors such as anesthesia type and fluid status have been suggested as potential contributors, the underlying cause remains unclear. With advancements in TAVI techniques, there has been a shift in anesthesia protocols from general anesthesia (GA) to monitored anesthesia care (MAC). This study aimed to assess the impact of GA and MAC on the increase in transvalvular gradients observed 24 h post-TAVI. Methods: A retrospective, single-center analysis was conducted on patients who underwent TAVI at our institution between 2011 and 2023 (n = 744, males = 421). The patients were divided into two groups: those who received GA (n = 201) and those who received MAC (n = 543). The GA group received either inhaled anesthetics, with or without propofol infusions, or propofol infusions at a rate of ≥100 mcg/kg/min. The MAC group received bolus doses and continuous infusions of dexmedetomidine. Transvalvular gradients were compared between immediate and 24 h post-procedure echocardiograms. Results: The average age of patients in the GA group (78 years [IQR 71–83]) was similar to that of the MAC group (77 years [IQR 71–83]). The GA group had a higher prevalence of comorbidities at baseline. Both groups exhibited stable, normotensive blood pressure levels during the procedure, though the GA group required more vasopressors and intravenous fluid. The GA group showed a 24 h post-TAVI mean transvalvular gradient change of +5.1 mmHg [IQR 3–8.1], while the MAC group had a 24 h mean transvalvular gradient change of +5.8 mmHg [IQR 3.2–9], with no significant difference between the groups (p = 0.139). Conclusions: Despite the greater cardiovascular depressive effects and increased need for vasopressors and fluid resuscitation in the GA group, there was no significant difference in the increase in transvalvular gradients between the GA and MAC groups at 24 h post-TAVI. Further research is needed to fully understand the reasons behind the increase in gradients observed after TAVI. Full article

Background/Objectives: Obesity is a major risk factor for diabetes mellitus, a metabolic disease characterized by elevated fasting blood glucose and glycosylated hemoglobin levels. Predicting the percentage and absolute variations in key medical parameters based on weight changes can help patients stay motivated to lose weight and assist doctors in making informed lifestyle and treatment recommendations. This study aims to assess the extent to which weight variation influences the absolute and percentage changes in various clinical parameters. Methods: The dataset includes medical records from patients in Bucharest hospitals, collected between 2012 and 2016. Several machine learning models, namely linear regression, polynomial regression, Gradient Boosting, and Extreme Gradient Boosting, were employed to predict changes in medical parameters as a function of body weight variation. Model performance was evaluated using Mean Squared Error, Mean Absolute Error, and R² score. Results: Almost all models demonstrated promising predictive performance. Quantitative predictions were made for each parameter, highlighting the relationship between weight loss and improvements in clinical indicators. Conclusions: Weight loss led to significant improvements in dysglycemia, dyslipidemia, inflammation, uric acid levels, liver enzymes, thyroid hormones, and blood pressure, with reductions ranging from 5% to 30%, depending on the parameter. Full article

Vascular steal refers to the diversion of blood flow between collateral vessels that share a common inflow restricted by arterial stenosis. Blood is diverted from the high-pressure to the low-pressure, low-resistance system. Vascular steal is associated with anatomical bypass or vasodilation in the collateral network and is called “the arterial steal”. However, we have demonstrated that in the presence of an outflow gradient (e.g., intra-extracranial), blood is shunted to a lower pressure system, a phenomenon we term “venous steal”. Using Thevenin’s equivalent, we generalized the concept of venous steal to apply it to any region of the vascular system with increased outflow pressure. Both arterial steal, caused by increased collateral network conductivity, and venous steal, resulting from lower collateral outflow pressure, reduce compartment perfusion. This occurs indirectly by increasing flow and the pressure gradient across the arterial stenosis, lowering the segmental compartment perfusion pressure—the difference between post-stenotic (inflow) and compartmental (outflow) pressures. Venous steal diverts blood flow from compartments with elevated pressure, such as intracranial, subendocardial, the ischemic core, and regions of focal edema due to inflammation, trauma, or external compression. In shock and low-flow states, it contributes to regional blood flow maldistribution. Treatment of venous steal addresses inflow stenosis, increased compartmental pressure and systemic loading conditions (arterial and venous pressure) to reverse venous steal malperfusion in the ischemic regions. Full article

Background: Recent studies suggest that any degree of patient-prosthesis mismatch (PPM) increases morbidity and mortality after surgical aortic valve replacement (SAVR). We used computational fluid dynamics simulations to test the influence of prosthesis size and physical activity after SAVR. Methods: In 10 patients with aortic valve stenosis, virtual SAVR was performed. Left ventricular outflow tract stroke volume and flow direction information (4D Flow) were used, and an increase in stroke volume of 25% was chosen for simulating physical activity. Pressure gradients (DP max) across the aortic valve and blood flow profiles in the ascending aorta were calculated and predicted for three different valve sizes at rest and under stress in every patient. Results: Gradients across the aortic valve were significantly lower using larger valves; however, they were not normalized after SAVR (DP max [mmHg] norm/smaller/reference/larger valve = 6/14/12/9 mmHg, <0.01 compared to norm). Physical activity simulation increased DP max in all patients and across all valve sizes (DP max [mmHg] rest versus stress for the smaller/reference/larger valve = 14 vs. 23, 12 vs. 18, 9 vs. 14). Blood flow profiles did not normalize after SAVR and remained unaffected by physical activity. Gradients differed between mild and moderate stenosis between different therapy options and even showed moderate to severe stenosis under simulated physical activity. Conclusions: Prosthesis size and physical activity simulation have a significant influence on gradients across the aortic valve. Virtual therapy planning using patient-specific data might help to improve outcomes after SAVR in the future. Full article

Background: Pressure injuries (PIs) are increasing worldwide, and there has been no significant improvement in preventing them. Traditional assessment tools are widely used to identify a patient at risk of developing a PI. This study aims to construct a novel fused multi-channel prediction model of PIs in adult hospitalized patients using machine learning algorithms (MLAs). Methods: A multi-phase quantitative approach involving a case–control experimental design was used. A first-hand dataset was collected retrospectively between March/2022 and August/2023 from the electronic medical records of three hospitals in Palestine. Results: The total number of patients was 49,500. A balanced dataset was utilized with a total number of 1110 patients (80% training and 20% testing). The models that were developed utilized eight MLAs, including linear regression and support vector regression (SVR), logistic regression (LR), random forest (RF), gradient boosting (GB), K-nearest neighbor (KNN), decision tree (DT), and extreme gradient boosting (XG boosting) and validated with five-fold cross-validation techniques. The best model was RF, for which the accuracy was 0.962, precision was 0.942, recall was 0.922, F1 was 0.931, area under curve (AUC) was 0.922, false positive rate (FPR) was 0.155, and true positive rate (TPR) was 0.782. Conclusions: The predictive factors were age, moisture, activity, length of stay (LOS), systolic blood pressure (BP), and albumin. A novel fused multi-channel prediction model of pressure injury was developed from different datasets. Full article

This paper presents a method to improve confidence-interval (CI) estimation using individual uncertainty measures and weighted feature decisions for cuff-less blood-pressure (BP) measurement. We obtained uncertainty using Gaussian process regression (GPR). The CI obtained from the GPR model is computed using the distribution of BP estimates, which provides relatively wide CIs. Thus, we proposed a method to obtain improved CIs for individual subjects by applying bootstrap and uncertainty methods using the cuff-less BP estimates of each subject obtained through GPR. This study also introduced a novel method to estimate cuff-less BP with high fidelity by determining highly weighted features using weighted feature decisions. The standard deviation of the proposed method’s mean error is 2.94 mmHg and 1.50 mmHg for systolic blood pressure (SBP) and (DBP), respectively. The mean absolute error results were obtained by weighted feature determination combining GPR and gradient boosting algorithms (GBA) for SBP (1.46 mmHg) and DBP (0.69 mmHg). The study confirmed that the BP estimates were within the CI based on the test samples of almost all subjects. The weighted feature decisions combining GPR and GBA were more accurate and reliable for cuff-less BP estimation. Full article

Objective: Cardiopulmonary bypass (CPB) via the right axillary artery (RAA) has become an alternative perfusion strategy, especially in complex aortic procedures. This study delineates our technique and outcome with direct axillary cannulation utilizing the Seldinger technique, which we adopted as the standard perfusion strategy in the sternum-sparing minimally invasive total coronary revascularization via left anterior thoracotomy (TCRAT) using CPB. Methods: From November 2019 to December 2023, a total of 413 consecutive patients underwent nonemergent isolated coronary artery bypass grafting (CABG) via left anterior minithoracotomy on CPB with peripheral cannulation via the RAA and cardioplegic cardiac arrest, using this technique as a default strategy in the daily routine. All patients had multivessel coronary artery disease. The primary outcome was intraoperative cannulation-related complications (bleeding, revision, ischemia, wound healing complications). The secondary outcome was cannulation-related events during follow-up (blood pressure differences, incidence of brachial plexus injury, clinical signs of circulatory problems of arm and hand, re-interventions). Mean midterm follow-up was 18.7 ± 12.3 [1.1–51.2] months. During follow-up, 16 patients died. Overall, a total of 397 patients (344 male; 67.6 ± 9.7 [32–88]) were included for follow-up (100%). Results: The RAA was successfully cannulated in 100% of patients. A cannula size of 16 Fr was used in 34.6%, 18 Fr in 63.9% and 20 Fr in 1.5% of all patients. There was no intraoperative bleeding complication. In two patients, intraoperative revision of the RAA was required, necessitating a venous patch repair. At follow-up, there were no differences between the systolic and diastolic blood pressure or the pressure gradients between the right and left arm. Transient numbness of the right hand was observed in two patients. Permanent numbness was not observed. No patient needed further intervention or surgical revision of the RAA. Conclusions: The right axillary cannulation is feasible and safe in terms of vascular injury and brachial plexus injury with excellent in-hospital and follow-up outcome. Full article

Background/Objectives: For effective exercise prescription for patients with cardiovascular disease, it is important to determine the target heart rate at the level of the anaerobic threshold (AT-HR). The AT-HR is mainly determined by cardiopulmonary exercise testing (CPET). The aim of this study is to develop a machine learning (ML) model to predict the AT-HR solely from non-exercise clinical features. Methods: From consecutive 21,482 cases of CPET between 2 February 2008 and 1 December 2021, an appropriate subset was selected to train our ML model. Data consisted of 78 features, including age, sex, anthropometry, clinical diagnosis, cardiovascular risk factors, vital signs, blood tests, and echocardiography. We predicted the AT-HR using a ML method called gradient boosting, along with a rank of each feature in terms of its contribution to AT-HR prediction. The accuracy was evaluated by comparing the predicted AT-HR with the target HRs from guideline-recommended equations in terms of the mean absolute error (MAE). Results: A total of 8228 participants included healthy individuals and patients with cardiovascular disease and were 62 ± 15 years in mean age (69% male). The MAE of the AT-HR by the ML-based model was 7.7 ± 0.2 bpm, which was significantly smaller than those of the guideline-recommended equations; the results using Karvonen formulas with the coefficients 0.7 and 0.4 were 34.5 ± 0.3 bpm and 11.9 ± 0.2 bpm, respectively, and the results using simpler formulas, rest HR + 10 and +20 bpm, were 15.9 ± 0.3 and 9.7 ± 0.2 bpm, respectively. The feature ranking method revealed that the features that make a significant contribution to AT-HR prediction include the resting heart rate, age, N-terminal pro-brain natriuretic peptide (NT-proBNP), resting systolic blood pressure, highly sensitive C-reactive protein (hsCRP), cardiovascular disease diagnosis, and β-blockers, in that order. Prediction accuracy with the top 10 to 20 features was comparable to that with all features. Conclusions: An accurate prediction model of the AT-HR from non-exercise clinical features was proposed. We expect that it will facilitate performing cardiac rehabilitation. The feature selection technique newly unveiled some major determinants of AT-HR, such as NT-proBNP and hsCRP. Full article

Background/Objectives: Pulmonary embolism (PE) is a potentially serious condition characterized by the blockage of blood vessels in the lungs, often presenting significant diagnostic challenges due to its non-specific symptoms. This study aimed to evaluate the utility of the alveolar–arterial (A-a) oxygen gradient as a diagnostic tool for PE, hypothesizing that it could enhance early detection when combined with other clinical markers. Methods: We retrospectively analyzed 168 patients at the University Hospital Center Split. This study correlated A-a gradients with PE confirmed by CT pulmonary angiography. Key clinical and biochemical markers, including heart rate, C-reactive protein (CRP), pro-brain natriuretic peptide (NT-proBNP), D-dimer, high-sensitivity troponin (hs-troponin), and arterial oxygen pressure (PaO₂), were assessed. Results: Our findings revealed that patients with PE had significantly higher A-a gradients than those without PE. The observed-to-expected ratio for the A-a gradient was notably increased in the PE group. Additionally, patients with PE exhibited elevated heart rate, CRP, NT-proBNP, D-dimer, and hs-troponin levels, while PaO₂ levels were notably lower. Conclusions: This study demonstrates that an elevated A-a gradient reflects the severity of gas exchange impairment in PE. The results suggest that early diagnosis of PE may be improved by incorporating A-a gradient analysis alongside other clinical markers, potentially leading to more effective and timely interventions. Full article

Background/Objectives: Non-invasive ventilation (NIV) has emerged as a possible first-step treatment to avoid invasive intubation in pediatric intensive care units (PICUs) due to its advantages in reducing intubation-associated risks. However, the timely identification of NIV failure is crucial to prevent adverse outcomes. This study aims to identify predictors of first-attempt NIV failure in PICU patients by testing various machine learning techniques and comparing their predictive abilities. Methods: Data were sourced from the TIPNet registry, which comprised patients admitted to 23 Italian Paediatric Intensive Care Units (PICUs). We selected patients between January 2010 and January 2024 who received non-invasive ventilation (NIV) as their initial approach to respiratory support. The study aimed to develop a predictive model for NIV failure, selecting the best Machine Learning technique, including Generalized Linear Models, Random Forest, Extreme Gradient Boosting, and Neural Networks. Additionally, an ensemble approach was implemented. Model performances were measured using sensitivity, specificity, AUROC, and predictive values. Moreover, the model calibration was evaluated. Results: Out of 43,794 records, 1861 admissions met the inclusion criteria, with 678 complete cases and 97 NIV failures. The RF model demonstrated the highest AUROC and sensitivity equal to 0.83 (0.64, 0.94). Base excess, weight, age, systolic blood pressure, and fraction of inspired oxygen were identified as the most predictive features. A check for model calibration ensured the model’s reliability in predicting NIV failure probabilities. Conclusions: This study identified highly sensitive models for predicting NIV failure in PICU patients, with RF as a robust option. Full article

Designing scaffolds similar to the structure of trabecular bone requires specialised algorithms. Existing scaffold designs for bone tissue engineering have repeated patterns that do not replicate the random stochastic porous structure of the internal architecture of bones. In this research, the Voronoi tessellation method is applied to create random porous biomimetic structures. A volume mesh created from the shape of a Zygoma fracture acts as a boundary for the generation of random seed points by point spacing to create Voronoi cells and Voronoi diagrams. The Voronoi lattices were obtained by adding strut thickness to the Voronoi diagrams. Gradient Voronoi scaffolds of pore sizes (19.8 µm to 923 µm) similar to the structure of the trabecular bone were designed. A Finite Element Method-based computational fluid dynamics (CFD) simulation was performed on all designed Voronoi scaffolds to predict the pressure drops and permeability of non-Newtonian blood flow behaviour using the power law material model. The predicted permeability (0.33 × 10⁻⁹ m² to 2.17 × 10⁻⁹ m²) values of the Voronoi scaffolds from the CFD simulation are comparable with the permeability of scaffolds and bone specimens from other research works. Full article