A Review of Machine Learning’s Role in Cardiovascular Disease Prediction: Recent Advances and Future Challenges

Algorithms, Год журнала: 2024, Номер 17(2), С. 78 - 78

Опубликована: Фев. 13, 2024

Cardiovascular disease is the leading cause of global mortality and responsible for millions deaths annually. The rate overall consequences cardiac can be reduced with early detection. However, conventional diagnostic methods encounter various challenges, including delayed treatment misdiagnoses, which impede course raise healthcare costs. application artificial intelligence (AI) techniques, especially machine learning (ML) algorithms, offers a promising pathway to address these challenges. This paper emphasizes central role in health focuses on precise cardiovascular prediction. In particular, this driven by urgent need fully utilize potential enhance light continued progress growing public implications disease, aims offer comprehensive analysis topic. review encompasses wide range topics, types significance learning, feature selection, evaluation models, data collection & preprocessing, metrics prediction, recent trends suggestion future works. addition, holistic view learning’s prediction health. We believe that our will contribute significantly existing body knowledge essential area.

Язык: Английский

---

Multilayer Perceptron Neural Network with Arithmetic Optimization Algorithm-Based Feature Selection for Cardiovascular Disease Prediction

Machine Learning and Knowledge Extraction, Год журнала: 2024, Номер 6(2), С. 987 - 1008

Опубликована: Май 5, 2024

In the healthcare field, diagnosing disease is most concerning issue. Various diseases including cardiovascular (CVDs) significantly influence illness or death. On other hand, early and precise diagnosis of CVDs can decrease chances death, resulting in a better healthier life for patients. Researchers have used traditional machine learning (ML) techniques CVD prediction classification. However, many them are inaccurate time-consuming due to unavailability quality data imbalanced samples, inefficient preprocessing, existing selection criteria. These factors lead an overfitting bias issue towards certain class label model. Therefore, intelligent system needed which accurately diagnose CVDs. We proposed automated ML model various kinds Our consists multiple steps. Firstly, benchmark dataset preprocessed using filter techniques. Secondly, novel arithmetic optimization algorithm implemented as feature technique select best subset features that accuracy Thirdly, classification task multilayer perceptron neural network classify instances into two labels, determining whether they not. The trained on then tested validated. Furthermore, comparative analysis model, performance evaluation metrics calculated overall accuracy, precision, recall, F1-score. As result, it has been observed achieve 88.89% highest comparison with

Язык: Английский

---

Performance discrepancy mitigation in heart disease prediction for multisensory inter-datasets

PeerJ Computer Science, Год журнала: 2024, Номер 10, С. e1917 - e1917

Опубликована: Март 18, 2024

Heart disease is one of the primary causes morbidity and death worldwide. Millions people have had heart attacks every year, only early-stage predictions can help to reduce number. Researchers are working on designing developing prediction systems using different advanced technologies, machine learning (ML) them. Almost all existing ML-based works consider same dataset (intra-dataset) for training validation their method. In particular, they do not inter-dataset performance checks, where datasets used in testing phases. setup, ML models show a poor named discrepancy problem. This work focuses mitigating problem by considering five available combined form. All potential mode combinations systematically executed assess discrepancies before after applying proposed methods. Imbalance data handling SMOTE-Tomek, feature selection random forest (RF), extraction principle component analysis (PCA) with long preprocessing pipeline mitigate The builds missing value RF regression, log transformation, outlier removal, normalization, balancing that convert more ML-centric. Support vector machine, K-nearest neighbors, decision tree, RF, eXtreme Gradient Boosting, Gaussian naive Bayes, logistic multilayer perceptron as classifiers. Experimental results classification produce better than other combination strategies both single- setups. certain configurations individual datasets, demonstrates 100% accuracy 96% during phase an exhibiting commendable precision, recall, F1 score, specificity, AUC score. indicate effective technique has improve model without necessitating development intricate models. Addressing introduces novel research avenue, enabling amalgamation identical features from various construct comprehensive global within specific domain.

Язык: Английский

---

Heart Disease Prediction Using Novel Quine McCluskey Binary Classifier (QMBC)

IEEE Access, Год журнала: 2023, Номер 11, С. 64324 - 64347

Опубликована: Янв. 1, 2023

Cardiovascular disease is the primary reason for mortality worldwide, responsible around a third of all deaths. To assist medical professionals in quickly identifying and diagnosing patients, numerous machine learning data mining techniques are utilized to predict disease. Many researchers have developed various models boost efficiency these predictions. Feature selection extraction remove unnecessary features from dataset, thereby reducing computation time increasing models. In this study, we introduce new ensemble Quine McCluskey Binary Classifier (QMBC) technique patients diagnosed with some form heart those who not diagnosed. The QMBC model utilizes an seven models, including logistic regression, decision tree, random forest, K-nearest neighbour, naive bayes, support vector machine, multilayer perceptron, performs exceptionally well on binary class datasets. We employ feature accelerate prediction process. utilize Chi-Square ANOVA approaches identify top 10 create subset dataset. then apply Principal Component Analysis 9 prime components. obtain Minimum Boolean expression target feature. results ( x ₀ , xmlns:xlink="http://www.w3.org/1999/xlink">1 xmlns:xlink="http://www.w3.org/1999/xlink">2 ..., xmlns:xlink="http://www.w3.org/1999/xlink">6 ) considered independent features, while attribute dependent. combine projected outcomes ML foaming utilizing minimum equation built 80:20 train-to-test ratio. Our proposed surpasses current state-of-the-art previously suggested methods put forward by researchers.

Язык: Английский

---

Injury Patterns and Impact on Performance in the NBA League Using Sports Analytics

Computation, Год журнала: 2024, Номер 12(2), С. 36 - 36

Опубликована: Фев. 16, 2024

This research paper examines Sports Analytics, focusing on injury patterns in the National Basketball Association (NBA) and their impact players’ performance. It employs a unique dataset to identify common NBA injuries, determine most affected anatomical areas, analyze how these injuries influence post-recovery study’s novelty lies its integrative approach that combines data with performance metrics salary data, providing new insights into relationship between economic on-court investigates periodicity seasonality of seeking related time external factors. Additionally, it effect specific per-match analytics performance, offering perspectives implications rehabilitation for player contributes significantly sports analytics, assisting coaches, medicine professionals, team management developing prevention strategies, optimizing rotations, creating targeted plans. Its findings illuminate interplay salaries, NBA, aiming enhance welfare league’s overall competitiveness. With comprehensive sophisticated analysis, this offers unprecedented dynamics long-term effects athletes.

Язык: Английский

---

TPTM-HANN-GA: A Novel Hyperparameter Optimization Framework Integrating the Taguchi Method, an Artificial Neural Network, and a Genetic Algorithm for the Precise Prediction of Cardiovascular Disease Risk

Mathematics, Год журнала: 2024, Номер 12(9), С. 1303 - 1303

Опубликована: Апрель 25, 2024

The timely and precise prediction of cardiovascular disease (CVD) risk is essential for effective prevention intervention. This study proposes a novel framework that integrates the two-phase Taguchi method (TPTM), hyperparameter artificial neural network (HANN), genetic algorithm (GA) called TPTM-HANN-GA. efficiently optimizes hyperparameters an (ANN) model during training stage, significantly enhancing accuracy risk. proposed TPTM-HANN-GA requires far fewer experiments than traditional grid search, making it highly suitable application in resource-constrained, low-power computers, edge intelligence (edge AI) devices. Furthermore, successfully identified optimal configurations ANN model’s hyperparameters, resulting hidden layer 4 nodes, tanh activation function, SGD optimizer, learning rate 0.23425849, momentum 0.75462782, seven nodes. optimized achieves 74.25% predicting disease, which exceeds existing state-of-the-art GA-ANN TSTO-ANN models. enables personalized CVD to be conducted on computers edge-AI devices, achieving goal point-of-care testing (POCT) empowering individuals manage their heart health effectively.

Язык: Английский

---

An offbeat bolstered swarm integrated ensemble learning (BSEL) model for heart disease diagnosis and classification

Applied Soft Computing, Год журнала: 2024, Номер 154, С. 111273 - 111273

Опубликована: Янв. 23, 2024

Язык: Английский

---

Predicting incident cardio-metabolic disease among persons with and without depressive and anxiety disorders: a machine learning approach

Social Psychiatry and Psychiatric Epidemiology, Год журнала: 2025, Номер unknown

Опубликована: Фев. 18, 2025

Язык: Английский

---

Advanced Hybrid Machine Learning Model for Accurate Detection of Cardiovascular Disease

International Journal of Computational Intelligence Systems, Год журнала: 2025, Номер 18(1)

Опубликована: Март 6, 2025

Cardiovascular disease (CVD) is one of the foremost reasons behind death people worldwide. Prevention and early diagnosis are only ways to control its progression onset. Thus, there an urgent need for a detection model comprising intelligent technologies, including Machine Learning (ML) deep learning, predict future state individual suffering from cardiovascular by effectively analyzing patient data. This study aims propose hybrid that provides insight into data under consideration enhance accuracy detecting disease. current research proposes four stages. In first stage proposed model, imbalance problem solved using sampling technique named Synthetic Minority Oversampling Technique-Edited Nearest Neighbors Rule. second stage, Chi-square applied as feature selection method select highly relevant features records 1190 with 11 clinical features, curated combining 5 most popular datasets, Long Beach VA, Hungarian, Switzerland, Statlog (Heart). third preprocessed dataset passed stacking ensemble three base learners: Random Forest Tree (RFT), K-Nearest Neighbor (K-NN), AdaBoost classifier meta-learner: Logistic Regression (LR), optimized Grid Search Cross-Validation (GSCV) optimization approach, whose performance evaluated against classifier. fourth in terms accuracy, sensitivity, specificity, F1 score, ROC_AUC score.. The comparative results prove scored highest 97.8%, 96.15% 96.75% specificity 98.6% score when compared existing techniques models after applying SMOTE–ENN (for balancing) selection) methods efficient implementation demonstrate suggested may accurately identify among patients. It facilitates application robust treatment strategies.

Язык: Английский

---

Enhancing cardiovascular risk prediction: the role of wall viscoelasticity in machine learning models

Computer Methods in Biomechanics & Biomedical Engineering, Год журнала: 2025, Номер unknown, С. 1 - 12

Опубликована: Март 10, 2025

This study aims to evaluate the significance of wall viscoelasticity in enhancing cardiovascular disease (CVD) risk prediction. We collected data on ten patient features, categorized into demographics (age, gender, blood pressure, smoking history), lab (HDL, LDL, glucose levels), and mechanics (Peterson's modulus, stiffness parameter, energy dissipation ratio). Outcome variables were classified as low or high CVD based total plaque area computed from carotid ultrasound images. employed eight machine learning classifiers conducted a comparative analysis feature importance. Incorporating mechanical attributes significantly improved predictive accuracies for most models. The Random Forest Bagging Method (RFBM) showed best performance, achieving an accuracy 93.0% AUC 0.98 with all 10 features. Including either elastic viscous features alongside conventional enhanced prediction For tree-based bagging models (DTBM RFBM), including (energy ratio) resulted greater improvements compared underscores significant impact integrating viscosity highlights potential combining both characteristics achieve more accurate assessment.

Язык: Английский

---