Heart Failure

1. Problem Statement.

Heart Failure defines as the failure of the heart to push adequate blood according to body requirements. The leading cause of Heart Failure is contraction or blocking of the coronary infarction. Coronary infarction is the arteries that carry blood to the heart. Breath difficulty, inflamed feet, and sickness in the body are the common symptoms of heart failure.

Proper diagnosis of heart failure considered a barrier because of the deficiency of appropriate symptomatic equipment and medical specialists. Conventional methods based on various medical tests have used for the diagnosis of heart failure recommended by physicians. Among them, an angiography is a vital tool for heart failure diagnosis. An angiography is a form of determination that has been used to diagnose heart disease and is considered a promising approach to heart failure. But it does have some constraints, such as:

·         significant expenses

• Side-effects related health
• High level of technical expertise is required

2. Objectives (To be attained).
3. The goal of this research is to improve the accuracy of prediction of heart failure and reduce the health risks by diagnosing the disease at early stages.
4. It helps cardiologists for decision making for the treatment of disease.

• It is eventually leading to cost savings and time savings during the treatment of heart failure.

1. Reduce the death rate

The ultimate goal of this research is to propose a method for the improvement and enhancement of accuracy of prediction of heart failure with deep learning techniques.

3. 3. Literature Survey

3. ALI et al.[1] in 2019 proposed the expert system based on two vector machine models to make the treatment of cardiac failure easier. The fist SVM model eliminated irrelevant features, and the second used as the basis for prediction. The hybrid grid search algorithm had used to optimize both the models. It also noted that the accuracy of the SVM model is improved by up to 3.3 % by the proposed model. Thus, As for the complexity of time, the proposed method is better and efficient. Hence, based on the experimental results that are obtained from the data set of Heart Failure, it is found that the proposed expert system can improve the decision-making process of the physicians.

   Don't use plagiarised sources.Get your custom essay just from $11/page

   GET CUSTOM PAPER

Kumar, PM et al.[2] developed an IoT-based architecture that was used to address the data and also defined important prescription parameters obtaining heart diseases. ROC analysis was used to show imminent heart disease. A positive correlation between heart disease was found in blood glucose. Moreover, the parameters which negatively correlated with heart disease are breathing speed, Heartbeat, strain, and temperature of the body. The obtained experimental results proved that in the indication of the cardiac condition, breathing rate around time and heartbeat is significant. Similarly, for cardiac problem indication, strain pulse range, and temperature of the body are also reflected as a substantial variation.

Raghavendra Kumar et al. [3] identified and compared the many techniques of machine learning using Random Forest (RF), Support Vector Machine (SVM), XG Boost and the Artificial Neural Network (ANN) reveal an F1 score, recall, precision to conclude the accurate and more efficient results. Existing benchmark models are used to compare the results, and significant improvement noticed in the prediction of heart disease in patients.

Arabasadi et al. [4] A highly accurate hybrid method was proposed by Arabasadi et al. for the recognition of disease of the arteries. The performance of the neural network increased by the proposed method, approximately up to 10%, by improving its initial weights with the help of a genetic algorithm. The genetic algorithm recommends better weights for neural networks. A new hybrid method was proposed in this article to increase the performance of the neural network.  Categorically, using this method, without angiography, CAD can be detected, which can help to reduce high costs and substantial side effects.

Mihaela Porumb et al.[5] Mihaela Porumb et al. proposed a novel based method. For Heart Failure recognition using a machine learning algorithm. Raw ECG signals were used in this model in place of HRV features and produced well-known accuracy in detecting CHF. The CNN models obtained considerably high scores in terms of performance. Also, this is an initial analysis in which the AML methods used to disclose specific structural properties of those electrocardiogram beats that are essential to identify Heart Failure successfully.

Mold Abdar et al.[6] Initially, the diagnosis of the cardiac artery disease approach presented by Moloud Abdar et al., this approach can increase the medical determination cycle. A machine-based detection system is the main contribution of this study to forecasting cardiac artery disease through an improved and efficient outcome by comparison with standard machine learning methods. Experimentation includes a normalization technique used to carry out data pre-processing. Moreover, by coupling, the genetic algorithm and particle swarm optimization method have been applied in two configurations. First, only the genetic algorithm and particle swarm optimization were used to optimize the classifiers parameters and second similar genetic algorithm, and particle swarm optimization founded components chosen. This work has shown that the prediction mechanism can be improved through the optimization approach.

Abdel-Basset et al.[7] For detection and monitoring of patients with Heart Failures Current decision-making model based on IoT was proposed by Abdel-Basset et al. portable device interact was used to catch the community interactions and indications of the human physique. After the personal data obtained, information as well as indicators, consumers classified within infected or uninfected people. In this work, a hybrid methodology proposed to develop a decision support system for the accurate forecast of risk of Heart Failure in patients.

Malav et al.[8] An predictive analysis was carried out by Malav et al. on UCI Heart Disease Data Set. In this analysis, K-means and ANN data mining techniques used.  Medical data is a blending of blurred and crisp values. Based on their properties, the information has classified. This classification executed by forming a model using the Artificial Neural Network and K-means algorithm. The proposed model focused on organizing the data according to cardiovascular diseases better to have a more reliable diagnosis.

Manogaran et al.[9] A deep learning approach for the diagnosis of heart disease was proposed by Manogaran et al. Parameters were divided among cardiac disease patients and healthy persons by using the MKL method. To distinguish healthily and patients with heart disease, the result obtained from the MKL method is given to the ANFIS classificatory. Different available machine learning approaches compared with the suggested model.

Mohan et al.[10] Identified the collection of raw heart knowledge that will assist in the sustainable redemption of human lives and the rapid identification of cardiac abnormalities. T Machine learning techniques were used, which provided discrimination to cardiac disease. The developed hybrid methodology was recycled, which combined with the characteristics of the random forest and linear method. The prediction model for heart disease provided an improved performance level with an accuracy level of 88.7%.

4. Methodological Approach

1. a) Brief Description of how to carry out the research

Already applied methods and techniques were firstly studied to classify the expert system to propose a suitable solution for effective prediction of Heart Failure. Identify the issues regarding previously proposed approaches. Then draw out a suitable solution for the given problem which can satisfy our objectives so that we can suggest a method or technique for the prediction of Heart Failure.

1. Experimentation

We will use or consider deep learning techniques and their results as compared with related work algorithms for cross-validation.

1. c) Experimental setup

• Consider related work algorithms and techniques for improving accuracy.
• Use deep learning techniques and cross-validate their results with existing methods.

1. d) Theoretical Studies

Many papers on this topic are studied so that we can easily understand the problems related to Heart Failure issues. Recent work on this topic will facilitate us to find out a feasible solution.

1. e) Expecting Results and Method for Analysis

The presumed results of the proposed approach will be betterment inaccuracy of effective prediction for the Heart Failure to make better the diagnosis process. We can examine the proposed method by comparing the result with already applied approaches and techniques for prediction of Heart Failure.

5. Utilization of research results

The goal of this research is for the improvement and enhancement of the accuracy of prediction of the Heart Failure and reduce the health risks by diagnosing the disease at early stages.

6. Research Time Table.

The steps of the expected time for each stage discussed in detail. The research methodology of the research shown in the table.