Revolutionizing COVID-19 Patient Identification: Multi-modal Data Analysis with Emphasis on CNN Algorithm

One of the most important aspects of managing and controlling the COVID-19 pandemic effectively worldwide is the timely and precise identification of patients. As a solution to this necessity, this study offers a novel and automated tool that precisely identifies COVID-19 patients by utilizing multi-modal data, such as pictures from CT scans, ECGs, and chest X-rays. The application process consists of two steps: a web-based questionnaire is completed first, and then medical photos must be sent for validation. Various deep learning and machine learning techniques, including CNN, KNN, Logistic Regression, Decision Tree, and Naive Bayes, were used to train and validate the model thoroughly. LSTM, InceptionV3, SVM, Resnet, and MobileNet were some of these models. The Convolutional Neural Network (CNN) algorithm has consistently proved to be the most effective approach. It showed remarkable recall, accuracy, and F-score, as well as a low false prediction rate. This work demonstrates the possibility of multi-modal data analysis and displays the extraordinary performance of the CNN algorithm in accurately and efficiently identifying COVID-19 patients. The research findings have great potential to transform patient identification, resource allocation, and, ultimately, the ongoing fight against COVID-19 since the virus continues to pose a threat to healthcare systems across the globe.