.png)
Nowadays, individuals are so preoccupied with their employment and other responsibilities that they neglect their health. Every day, more individuals become ill because of their stressful lifestyles and lack of awareness about their health. Furthermore, most individuals suffer from illnesses such as heart disease. Heart disease, often known as cardiovascular disease, encompasses a variety of heart-related disorders and has been the leading cause of mortality globally in recent decades. Cardiovascular diseases (CVDs) are heart and blood vessel problems that include heart attacks, strokes, heart failure, and other pathologies. Heart failure (HF) develops when the heart cannot pump enough blood to fulfil the body's demands. Patients' electronic medical records measure symptoms, physical attributes, and clinical laboratory test data, enabling biostatistics analysis to identify patterns and connections that might otherwise go undetected by medical practitioners (Chicco & Jurman, 2020).
According to the World Health Organisation (WHO), over 17.9 million people die each year from heart-related complications, including heart disease. Heart attacks and strokes account for more than four out of every five CVD fatalities (Islam et al., 2022). Unhealthy diets, a lack of physical exercise, alcohol misuse, and tobacco use are all significant risk factors for heart-related issues. As a result, individuals exhibit intermediate-risk factors such as high blood pressure, high blood glucose levels, abnormal blood lipids, overweight, and obesity. However, identifying patients at high risk of CVD early on and giving proper therapy can help to prevent unexpected and untimely deaths. More than three-quarters of all CVD fatalities occur in low- and middle-income countries (Islam et al., 2022). With an increasing population and number of heart disease patients, it is becoming more challenging to provide individuals with affordable diagnoses in developing countries such as Bangladesh, India, and some African nations, where proper screening procedures for patients with heart disease symptoms are still questionable due to financial crises and a lack of adequate and equitable health care equipment and facilities (Islam et al., 2022).
According to the European Society of Cardiology, roughly 26 million persons with HD have been diagnosed, with 3.6 million diagnosed each year. Most people in the United States suffer from heart disease (Li et al., 2020). A physician typically diagnoses HD by analysing the patient's medical history, physical examination report, and symptoms. However, the data produced from this diagnosis procedure do not accurately identify HD patients. Furthermore, analysing it is costly and computationally complex. Experts propose developing a non-invasive diagnostic system based on machine learning (ML) classifiers to address these difficulties (Li et al., 2020).
SYMPTOMS OF HEART DISEASE
A growing body of research indicates that depression is an additional modifiable risk factor for heart disease. For instance, one study discovered that taking psychological characteristics like sadness into account decreased the risk of myocardial infarction by almost 30% (Deschênes et al., 2019). The risk of cardiovascular illnesses was somewhat elevated by depression; interestingly, the increase in risk was comparable to that of diabetes and smoking. Additionally, research has shown a link between subsyndromal depression and a higher risk of heart disease. When combined, depression and sleep issues have been linked to an increased risk of heart disease.
Nonetheless, there is a connection between insomnia and depression symptoms. Short or lengthy sleep duration, as well as poor sleep quality, are frequently associated with depression. Depression and sleep issues probably aggravate a similar route, resulting in a significantly increased risk of heart disease in those with both illnesses. Previous studies have connected depression subtypes to the risk of health diseases such as diabetes, and depression with sleep issues is probably another subtype of depression that is highly associated with heart disease. Despite the tight links between sleep and depression, it is unclear whether these variables work together to enhance the risk of heart disease (Deschênes et al., 2019).
Machine learning can estimate patient survival based on data and identify the most essential aspects of their medical records. It identifies several risk factors for heart disease and emphasises the importance of accurate, trustworthy, and rational approaches to early diagnosis and disease management. Data mining is a widely used approach for processing massive amounts of data in the healthcare industry. Researchers use various data mining and machine learning approaches to examine large amounts of complicated medical data, assisting healthcare providers in predicting cardiac disease (Shah et al., 2020). Heart disease has a high death rate across the world and has become a health risk for many people. Early detection of heart illness can save many lives; nevertheless, recognising cardiovascular disorders such as heart attacks and coronary artery disease is a crucial problem presented by frequent clinical data analysis. Machine learning (ML) can provide an effective solution for decision-making and accurate prediction (Kavitha et al., 2021).
Furthermore, given the present facilities, it is too expensive for the public to seek a heart disease diagnosis. It is the primary cause of mortality in adults. Machine learning can identify who is more likely to be diagnosed with heart disease based on their medical history. It recognises who has any indications of heart illness, such as chest discomfort or high blood pressure. It can aid in detecting diseases with fewer medical tests and more effective therapies, allowing them to be treated appropriately (Islam et al., 2022).
An electrocardiogram (ECG) diagnoses cardiovascular disease. However, visually detecting long-term ECG abnormalities takes time and effort. With the introduction of machine learning (ML) applications in medicine, several researchers and practitioners discovered machine learning-based heart disease detection (MLBHDD) systems to be cost-effective and versatile. As a result, various research suggested MLBHDD based on various heart disease datasets (Islam et al., 2022).
An expert judgement system based on machine learning classifiers and artificial fuzzy logic successfully diagnoses HD, lowering the fatality rate. Various researchers used the Cleveland heart disease data collection to identify the problem of HD. Machine learning prediction models require accurate data for training and testing. The performance of a machine learning model can be improved by training and testing it on a balanced dataset (Li et al., 2020).
The medical business is making significant progress in employing machine learning techniques. The proposed study proposes a unique machine-learning technique for predicting cardiac disease. The proposed research makes use of the Cleveland heart disease dataset, as well as data mining techniques, including regression and classification. Machine learning techniques such as Random Forest and Decision Tree are used (Kavitha et al., 2021). The machine learning model employs a revolutionary method. In the implementation, three machine learning methods are random forest, decision tree, and hybrid model. The experimental findings reveal an accuracy level of 88.7% using the heart disease prediction model with the hybrid model. The interface collects user input parameters to forecast heart disease using a hybrid Decision Tree and Random Forest model (Kavitha et al., 2021).
One of the possible downsides of machine learning (ML) and deep learning-based solutions is that there often needs to be a better explanation of the models' behaviours during final forecasts. For example, a Deep Learning-based model may have numerous hidden layers, but it is sometimes impossible to determine how each layer contributes to the final prediction. Aside from that, the biased performance of ML algorithms towards the majority class is another possible issue. A majority class happens when a data set includes the highest value in one class relative to the other courses, also known as an unbalanced dataset (Islam et al., 2022). The high dimensionality of data is a typical issue in machine learning; the datasets we utilise include enormous amounts of data, often making it impossible to examine the data in 3D, a phenomenon known as the "curse of dimensionality." As a result, processing this data takes a significant amount of memory, and overfitting might occur as the data grows exponentially. Using the weighting characteristics makes it possible to reduce dataset redundancy, which in turn aids in lowering execution processing times. Various feature engineering and feature selection strategies can be used to exclude material that could be more important to reduce the dimensionality of the dataset (Bharti et al., 2021).
On the other hand, a proper machine-learning model is required for optimal outcomes. A successful machine learning model works well on data observed during training (otherwise, the model could learn the training data) and on unseen data. To test all classifiers on data and discover that they correctly predict 50% of the instances, effective cross-validation methodologies and performance assessment criteria are crucial for a model's training and testing on a dataset (Li et al., 2020). When applied to medical records, machine learning may be an excellent tool for predicting each patient's survival with heart failure symptoms and detecting the most critical clinical traits (or risk factors) that may lead to heart failure. Scientists may use machine learning for both clinical prediction and feature rating. Computational intelligence demonstrates predictive potential when applied to medical information or combined with imaging. Furthermore, deep learning and meta-analysis research applied to this topic have lately surfaced in the literature, enhancing the performance of human professionals but at a lower accuracy (0.75 versus 0.59) (Chicco & Jurman, 2020).
To your health,
Nwabekee
REFERENCE
Bharti, R., Khamparia, A., Shabaz, M., Dhiman, G., Pande, S., & Singh, P. (2021). Prediction of heart disease using a combination of machine learning and Deep Learning. Computational Intelligence and Neuroscience, 2021, 1–11. https://doi.org/10.1155/2021/8387680
Chicco, D., & Jurman, G. (2020). Machine learning can predict the survival of patients with heart failure from serum creatinine and ejection fraction alone. BMC Medical Informatics and Decision Making, 20(1). https://doi.org/10.1186/s12911-020-1023-5
Deschênes, S. S., Burns, R. J., Graham, E., & Schmitz, N. (2019). Depressive symptoms and sleep problems as risk factors for heart disease: A prospective community study. Epidemiology and Psychiatric Sciences, 29. https://doi.org/10.1017/s2045796019000441
Islam, A. K. M. M., Nahar, J., Dutta, A., Li, Y., Benhar, H., Fahimnia, B., Acharya, U. R., Rajesh, K. N. V. P. S., Yang, W., Sellami, A., Awad, A., Romdhane, T. F., Gan, D., Sharma, L. D., Wang, X., Marateb, H. R., Yang, P., Brunese, L., Ammar, A., … Verma, S. (2022, March 29). Machine learning-based heart disease diagnosis: A systematic literature review. Artificial Intelligence in Medicine. https://www.sciencedirect.com/science/article/abs/pii/S0933365722000549
Kavitha, M., Gnaneswar, G., Dinesh, R., Sai, Y. R., & Suraj, R. S. (2021). Heart disease prediction using Hybrid Machine Learning Model. 2021 6th International Conference on Inventive Computation Technologies (ICICT). https://doi.org/10.1109/icict50816.2021.9358597
Li, J. P., Haq, A. U., Din, S. U., Khan, J., Khan, A., & Saboor, A. (2020). Heart disease identification method using machine learning classification in E-Healthcare. IEEE Access, 8, 107562–107582. https://doi.org/10.1109/access.2020.3001149
Shah, D., Patel, S., & Bharti, S. K. (2020). Heart disease prediction using Machine Learning Techniques. SN Computer Science, 1(6). https://doi.org/10.1007/s42979-020-00365-y
Pretty educative❤️
Educative
Informative!