An Efficient Deep Neural Network for Cardiovascular disease prediction

Cardiovascular disease (CVDs) is forecasted and diagnosed in medical informatics using deep learning-based algorithms. These models can identify the optimum course of action, identify phenotypes for complex disorders, and recognize clinical signs. Using a big dataset of patient medical data, a deep learning model can be developed as a means of predicting CVDs. This work examined CVDs for early prediction using deep learning-based regression analysis on a dataset of 70,000 medical records from Kaggle that included age, symptoms, and CVD data. We suggest using long short-term memory-based deep neural networks in regression analysis to detect CVD early on. It has been shown that the overlapping symptoms of the diseases increase accuracy when the diseases are simulated in pairs. 95% of the forecasts for cardiovascular sickness are identified, according to the study's results, and 92% of those predictions correctly identified dyspnea. However, examining the three cases—cyanosis, dyspnea, and chest discomfort—could increase the accuracy to 93%. Weakness, emptiness, and exhaustion all had 92% percentages. Our proposed study illustrates combos with fever, combining Weakness and Fatigue, Dyspnea, Chest Pain, Cyanosis, Emptysis, and Pressure Pain in the Chest, with optimal accuracy levels approaching 95%. We demonstrate the effectiveness of our suggested strategy on a number of evaluation parameters.