• International Journal of Internet, Broadcasting and Communication
• /
• v.16 no.4
• /
• pp.290-300
• /
• 2024

This study presents a comprehensive evaluation of various machine learning models for predicting heart failure outcomes. Leveraging a data set of clinical records, the performance of Logistic Regression, Support Vector Machine (SVM), Random Forest, Soft Voting ensemble, and XGBoost models are rigorously assessed using multiple evaluation metrics, including accuracy, precision, recall, F1-score, and area under the receiver operating characteristic curve (AUC). The analysis reveals that the XGBoost model outperforms the other techniques across all metrics, exhibiting the highest AUC score, indicating superior discriminative ability in distinguishing between patients with and without heart failure. Furthermore, the study highlights the importance of feature importance analysis provided by XGBoost, offering valuable insights into the most influential predictors of heart failure, which can inform clinical decision-making and patient management strategies. The research also underscores the significance of balancing precision and recall, as reflected by the F1-score, in medical applications to minimize the consequences of false negatives.

• Journal of the Korean Society of Manufacturing Process Engineers
• /
• v.12 no.4
• /
• pp.22-28
• /
• 2013

It is important to begin left ventricular assist device (LVAD) treatment at appropriate time for heart failure patients who expect cardiac recovery after the therapy. In order to predict the optimal timing of LVAD implantation, we predicted pumping efficacy of LVAD according to the severity of heart failure theoretically. We used LVAD-implanted cardiovascular system model which consist of 8 Windkessel compartments for the simulation study. The time-varying compliance theory was used to simulate ventricular pumping function in the model. The ventricular systolic dysfunction was implemented by increasing the end-systolic ventricular compliance. Using the mathematical model, we predicted cardiac responses such as left ventricular peak pressure, cardiac output, ejection fraction, and stroke work according to the severity of ventricular systolic dysfunction under the treatments of continuous and pulsatile LVAD. Left ventricular peak pressure, which indicates the ventricular loading condition, decreased maximally at the 1st level heart-failure under pulsatile LVAD therapy and 2nd level heart-failure under continuous LVAD therapy. We conclude that optimal timing for pulsatile LVAD treatment is 1st level heart-failure and for continuous LVAD treatment is 2nd level heart-failure when considering LVAD treatment as "bridge to recovery".

• Korean Circulation Journal
• /
• v.53 no.9
• /
• pp.621-631
• /
• 2023

Background and Objectives: The morphology-voltage-P-wave duration (MVP) electrocardiography (ECG) risk score is a newly defined scoring system that has recently been used for atrial fibrillation (AF) prediction. The aim of this study was to evaluate the ability of the MVP ECG risk score to predict AF in patients with an implantable cardioverter defibrillator (ICD) and heart failure with reduced ejection fraction in long-term follow-up. Methods: The study used a single-center, and retrospective design. The study included 328 patients who underwent ICD implantation in our hospital between January 2010 and April 2021, diagnosed with heart failure. The patients were divided into low, intermediate and high-risk categories according to the MVP ECG risk scores. The long-term development of atrial fibrillation was compared among these 3 groups. Results: The low-risk group included 191 patients, the intermediate-risk group 114 patients, and the high-risk group 23 patients. The long-term AF development rate was 12.0% in the low-risk group, 21.9% in the intermediate risk group, and 78.3% in the high-risk group. Patients in the high-risk group were found to have 5.2 times higher rates of long-term AF occurrence compared to low-risk group. Conclusions: The MVP ECG risk score, which is an inexpensive, simple and easily accessible tool, was found to be a significant predictor of the development of AF in the long-term follow-up of patients with an ICD with heart failure with reduced ejection fraction. This risk score may be used to identify patients who require close follow-up for development and management of AF.

• The Korean journal of internal medicine
• /
• v.39 no.1
• /
• pp.77-85
• /
• 2024

Background/Aims: There may be many predictors of anticoagulation-related gastrointestinal bleeding (GIB), but until now, systematic reviews and assessments of the certainty of the evidence have not been published. We conducted a systematic review to identify all risk factors for anticoagulant-associated GIB to inform risk prediction in the management of anticoagulation-related GIB. Methods: A systematic review and meta-analysis were conducted to search PubMed, EMBASE, Web of Science, and Cochrane Library databases (from inception through January 21, 2022) using the following search terms: anticoagulants, heparin, warfarin, dabigatran, rivaroxaban, apixaban, DOACs, gastrointestinal hemorrhage, risk factors. According to inclusion and exclusion criteria, studies of risk factors for anticoagulation-related GIB were identified. Risk factors for anticoagulant-associated GIB were used as the outcome index of this review. Results: We included 34 studies in our analysis. For anticoagulant-associated GIB, moderate-certainty evidence showed a probable association with older age, kidney disease, concomitant use of aspirin, concomitant use of the antiplatelet agent, heart failure, myocardial infarction, hematochezia, renal failure, coronary artery disease, helicobacter pylori infection, social risk factors, alcohol use, smoking, anemia, history of sleep apnea, chronic obstructive pulmonary disease, international normalized ratio (INR), obesity et al. Some of these factors are not included in current GIB risk prediction models. such as anemia, co-administration of gemfibrozil, co-administration of verapamil or diltiazem, INR, heart failure, myocardial infarction, etc. Conclusions: The study found that anemia, co-administration of gemfibrozil, co-administration of verapamil or diltiazem, INR, heart failure, myocardial infarction et al. were associated with anticoagulation-related GIB, and these factors were not in the existing prediction models. This study informs risk prediction for anticoagulant-associated GIB, it also informs guidelines for GIB prevention and future research.

• International Journal of Computer Science & Network Security
• /
• v.21 no.6
• /
• pp.77-88
• /
• 2021

Heart Failure represents a critical pathological case that is challenging to predict and discover at an early age, with a notable increase in morbidity and mortality. Machine Learning and Neural Network techniques play a crucial role in predicting heart attacks, diseases and more. These techniques give valuable perspectives for clinicians who may then adjust their diagnosis for each individual patient. This paper evaluated neural network models for heart attacks predictions. Several online learning methods were investigated to automatically and accurately predict heart attacks. The UCI dataset was used in this work to train and evaluate First Order and Second Order Online Learning methods; namely Backpropagation, Delta bar Delta, Levenberg Marquardt and QuickProp learning methods. An optimizer technique was also used to minimize the random noise in the database. A regularization concept was employed to further improve the generalization of the model. Results show that a three layers' NN model with a Backpropagation algorithm and Nadam optimizer achieved a promising accuracy for the heart attach prediction tasks.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.8
• /
• pp.2016-2029
• /
• 2023

Accurate prediction of critical illness is significant for ensuring the lives and health of patients. The selection of indicators affects the real-time capability and accuracy of the prediction for critical illness. However, the diversity and complexity of these indicators make it difficult to find potential connections between them and critical illnesses. For the first time, this study proposes an indicator analysis model to extract key indicators from the preoperative and intraoperative clinical indicators and laboratory results of critical illnesses. In this study, preoperative and intraoperative data of heart failure and respiratory failure are used to verify the model. The proposed model processes the datum and extracts key indicators through four parts. To test the effectiveness of the proposed model, the key indicators are used to predict the two critical illnesses. The classifiers used in the prediction are light gradient boosting machine (LightGBM) and eXtreme Gradient Boosting (XGBoost). The predictive performance using key indicators is better than that using all indicators. In the prediction of heart failure, LightGBM and XGBoost have sensitivities of 0.889 and 0.892, and specificities of 0.939 and 0.937, respectively. For respiratory failure, LightGBM and XGBoost have sensitivities of 0.709 and 0.689, and specificity of 0.936 and 0.940, respectively. The proposed model can effectively analyze the correlation between indicators and postoperative critical illness. The analytical results make it possible to find the key indicators for postoperative critical illnesses. This model is meaningful to assist doctors in extracting key indicators in time and improving the reliability and efficiency of prediction.

• International Journal of Heart Failure
• /
• v.6 no.1
• /
• pp.11-19
• /
• 2024

The prevalence of heart failure (HF) is increasing, necessitating accurate diagnosis and tailored treatment. The accumulation of clinical information from patients with HF generates big data, which poses challenges for traditional analytical methods. To address this, big data approaches and artificial intelligence (AI) have been developed that can effectively predict future observations and outcomes, enabling precise diagnoses and personalized treatments of patients with HF. Machine learning (ML) is a subfield of AI that allows computers to analyze data, find patterns, and make predictions without explicit instructions. ML can be supervised, unsupervised, or semi-supervised. Deep learning is a branch of ML that uses artificial neural networks with multiple layers to find complex patterns. These AI technologies have shown significant potential in various aspects of HF research, including diagnosis, outcome prediction, classification of HF phenotypes, and optimization of treatment strategies. In addition, integrating multiple data sources, such as electrocardiography, electronic health records, and imaging data, can enhance the diagnostic accuracy of AI algorithms. Currently, wearable devices and remote monitoring aided by AI enable the earlier detection of HF and improved patient care. This review focuses on the rationale behind utilizing AI in HF and explores its various applications.

• Journal of Yeungnam Medical Science
• /
• v.38 no.4
• /
• pp.337-343
• /
• 2021

Background: Chromogranin A (CgA) levels have been reported to predict mortality in patients with heart failure. However, information on the prognostic value and clinical availability of CgA is limited. We compared the prognostic value of CgA to that of previously proven natriuretic peptide biomarkers in patients with acute heart failure. Methods: We retrospectively evaluated 272 patients (mean age, 68.5±15.6 years; 62.9% male) who underwent CgA test in the acute stage of heart failure hospitalization between June 2017 and June 2018. The median follow-up period was 348 days. Prognosis was assessed using the composite events of 1-year death and heart failure hospitalization. Results: In-hospital mortality rate during index admission was 7.0% (n=19). During the 1-year follow-up, a composite event rate was observed in 12.1% (n=33) of the patients. The areas under the receiver-operating characteristic curves for predicting 1-year adverse events were 0.737 and 0.697 for N-terminal pro-B-type natriuretic peptide (NT-proBNP) and CgA, respectively. During follow-up, patients with high CgA levels (>158 pmol/L) had worse outcomes than those with low CgA levels (≤158 pmol/L) (85.2% vs. 58.6%, p<0.001). When stratifying the patients into four subgroups based on CgA and NT-proBNP levels, patients with high NT-proBNP and high CgA had the worst outcome. CgA had an incremental prognostic value when added to the combination of NT-proBNP and clinically relevant risk factors. Conclusion: The prognostic power of CgA was comparable to that of NT-proBNP in patients with acute heart failure. The combination of CgA and NT-proBNP can improve prognosis prediction in these patients.

• Korean Circulation Journal
• /
• v.53 no.3
• /
• pp.170-184
• /
• 2023

Background and Objectives: Several cases involving severe right ventricular (RV) failure in advanced cancer patients have been found to be pulmonary tumor thrombotic microangiopathies (PTTMs). This study aimed to discover the nature of rapid RV failure syndrome with a suspected diagnosis of PTTM for better diagnosis, treatment, and prognosis prediction in clinical practice. Methods: From 2011 to 2021, all patients with clinically suspected PTTM were derived from the one tertiary cancer hospital with more than 2000 in-hospital bed. Results: A total of 28 cases of clinically suspected PTTM with one biopsy confirmed case were included. The most common cancer types were breast (9/28, 32%) and the most common tissue type was adenocarcinoma (22/26, 85%). The time interval from dyspnea New York Heart Association (NYHA) Grade 2, 3, 4 to death, thrombocytopenia to death, desaturation to death, admission to death, RV failure to death, cardiogenic shock to death were 33.5 days, 14.5 days, 7.4 days, 6.4 days, 6.1 days, 6.0 days, 3.8 days and 1.2 days, respectively. The NYHA Grade 4 to death time was 7 days longer in those who received chemotherapy (7.1 days vs. 13.8 days, p value=0.030). However, anticoagulation, vasopressors or intensive care could not change clinical course. Conclusions: Rapid RV failure syndrome with a suspected diagnosis of PTTM showed a rapid progressive course from symptom onset to death. Although chemotherapy was effective, increased life survival was negligible, and treatments other than chemotherapy did not help to improve the patient's prognosis.

• International Journal of Computer Science & Network Security
• /
• v.24 no.10
• /
• pp.1-16
• /
• 2024

Due to its complexity and high diagnosis and treatment costs, heart attack (HA) is the top cause of death globally. Heart failure's widespread effect and high morbidity and death rates make accurate and fast prognosis and diagnosis crucial. Due to the complexity of medical data, early and accurate prediction of HA is difficult. Healthcare providers must evaluate data quickly and accurately to intervene. This novel hybrid approach predicts HA using Long Short-Term Memory (LSTM) networks, Deep belief networks (DBNs) with attention mechanism, and robust data mining to fill this essential gap. HA is predicted using Kaggle, PhysioNet, and UCI datasets. Wearable sensor data, ECG signals, and demographic and clinical data provide a solid analytical base. To maintain consistency, ECG signals are normalized and segmented after thorough cleaning to remove missing values and noise. Feature extraction employs complex approaches like Principal Component Analysis (PCA) and Autoencoders to pick time-domain (MNN, SDNN, RMSSD, PNN50) and frequency-domain (PSD at VLF, LF, HF bands) characteristics. The hybrid model architecture uses LSTM networks for sequence learning and DBNs for feature representation and selection to create a robust and comprehensive prediction model. Accuracy, precision, recall, F1-score, and ROC-AUC are measured after cross-entropy loss and SGD optimization. The LSTM-DBN model outperforms predictive methods in accuracy, sensitivity, and specificity. The findings show that several data sources and powerful algorithms can improve heart attack predictions. The proposed architecture performed well on many datasets, with an accuracy rate of 96.00%, sensitivity of 98%, AUC of 0.98, and F1-score of 0.97. High performance proves this system's dependability. Moreover, the proposed approach is outperformed compared to state-of-the-art systems.