AIDS's Drugs Quantification and Surveillance Using Deep Learning

Abstract

Consumption of medicines for a particular disease can be an indicator of the spread of the disease, as the increase in the consumption of medicines implies an increase in the incidence of the disease. Acquired immunodeficiency syndrome (AIDS) is a chronic, potentially life-threatening condition caused by the human immunodeficiency virus (HIV). AIDS is one of the deadliest diseases in human life. Therefore, monitoring the spread of AIDS through analyzing the consumption of its drugs and determining the places where the drugs are consumed geographically is an urgent necessity and brings useful information in the health sector. The main idea behind this paper is to employ a new approach of using deep learning as the main stage to predict the quantities of AIDS's drugs. Additionally, in the second stage the spatial concept is exploited to state the spread position of that disease. The deep neural network is a fully automated network that consists of a preprocessing layer, normalization layer and prediction layer depending on the state utilization drugs dataset of the USA for five consecutive years. Based on the results of the prediction process, the second stage represents the consumption of AIDS's drugs and produces a spatial map representing the disease Surveillance map. The results of the prediction process using the deep neural network are compared with the results of the linear regression method, as indicated by previous research. The deep network has given superior results by obtaining a very low value of (MSE).The results were encouraging and promising to rely on deep learning. The creation of a spatial map illustrated the spread of the disease and gave a clear vision based on the consumption of drugs.

Keywords:Risk Management, process management etc.