Daily Platelet Count Prediction in Treating Dengue Patients Using Deep Learning Algorithm

With the impact of deep learning algorithms, health care has expanded beyond all recognition in recent years. Applications of artificial intelligence that make use of data are more prevalent in health care. Numerous societal health challenges are resolved with the help of these programs. However, creating these applications requires changing the data's original format to one that the system can understand. Additionally, it entails the use of proper algorithms for the situation at hand. This article discusses a deep learning technique for predicting platelet count in those patients with dengue. The most significant virus spread by mosquitoes to infect humans is dengue. Despite the fact that it typically presents as a self-limited febrile sickness, problems could develop after the fever wears off. The most significant complication is a systemic vascular leak syndrome, which can occasionally progress to a potentially fatal hypovolemic shock and is frequently accompanied by hemoconcentration and thrombocytopenia. Additionally, it also causes a drop in the platelet count, which causes excessive bleeding. We examined information from a retrospective observational study of patients from 8 to 90 years old who were admitted between 2014 and 2018 to the Father Muller Medical College Hospital in Mangalore City with clinically suspected dengue. All patients with dengue made up the analysis population. This prediction model was built using artificial neural network (ANN). With only a few more variables plus the platelet counts from the previous two days, this model can accurately predict the platelet count for the following day to within about 90% of accuracy. .