Key result
The proposed Integrated Deep Learning Model with Convolution Neural Network (IDLM_CNN) achieved disease prediction accuracies of 88% for bradycardia, 93% for tachycardia, and 98% for myocardia, outperforming existing models.
Why the study?
Although heart disease prediction has been well studied using various features and methods, the accuracy of predicting heart disease remains a questioning factor.
An integrated deep learning model combining lung images, diabetic, and clinical features demonstrated high accuracy in predicting specific cardiac conditions compared to traditional machine learning models.
Hypothesis-generating for CNN-based cardiac prediction; prospective validation required before clinical adoption.
The Phenomenon of heart disease prediction has been well studied. There exist numerous techniques exist in literature which uses different features and methods. However, the accuracy of predicting heart disease is still a questioning factor. Towards improving the performance of heart disease prediction an efficient Integrated Deep Learning Model with Convolution Neural Network (IDLM_CNN) is presented in this article. The model considers various features from different data sets of lungs, diabetic and clinical features. The integrated model extracts texture features from lung images in form of mass values. Similarly, the blood glucose, BMI and other diabetic features are extracted from diabetic data set. Also, lifestyle features like physical habits, food habits and smoking habits are extracted from clinical data sets. Such features extracted from various data sets are combined and trained with Convolution neural network to support the disease prediction. The method convolves the features of lungs and combines with other features to compute Disease Prone Weight (DPW) towards cardiac disease. Based on the value of DPW, the method predicts the possibility of heart disease. The proposed method increases the performance of disease prediction and reduces the false ratio.
No takes yet. Share an insight, caveat, or question.
Syed Anwar Hussainy Fazlur (2022) studied Heart disease (Bradycardia, Tachycardia, Myocardia) (n=100,000). Integrated Deep Learning Model with Convolution Neural Network (IDLM_CNN) vs. XGBSVM, RERF-ILM, and MDCNN models was evaluated on Disease prediction accuracy. The proposed Integrated Deep Learning Model with Convolution Neural Network (IDLM_CNN) achieved disease prediction accuracies of 88% for bradycardia, 93% for tachycardia, and 98% for myocardia, outperforming existing models.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: