An Empirical Survey on Early Health Condition Prediction based on Clinical Parameters

Healthcare experts need accurate estimates of the outcomes of various diseases that patients suffer from. In addition, time is another important feature that affects clinical choices for accurate predictions. When there is a sickness in the body, the human body offers symptoms in the form of discomfort, changes in bodily parameters, changes in habits, and a variety of other changes. These changes are frequently assessed in order to detect a specific type of sickness. For example, Fluctuations in the blood glucose level is used to identify diabetes, whereas variations in the electrocardiogram signals are used to detect heart disease and other cardiac diseases. Although these diseases are frequently linked to one another, developing a crossover parametric model for each of them can help to improve the precision with which they are recognized. For example, analyzing the influence of fluctuations in blood sugar levels can be connected to the identification of cancer because long-term glucose consumption weakens the body and consequently diminishes its immune system's effectiveness. This decline in immunity causes a reduction in red blood cells, making the body more susceptible to malignancy. Many cases of cross parameter illness detection have been investigated over the years, and the results have been promising. Here several systems are compared which perform cross-analysis of cancer, cardiovascular disease, and diabetes. It first tests different cross-body parameter links between these diseases, and then tests the classification efficiency of a several machine learning / deep learning algorithms for detection of diseases by using cross-body parameters tested earlier. When compared to non-cross parametric techniques, it has been discovered that the system increases the accuracy of detection of various disease. In this paper we have examined few papers according to their methods, styles, actions, and procedures.