Abstract

The sensitivity of the display screen to human touch enables the ability to interact with a computer by touching images or text on a touch screen. Extended use of touch displays, particularly in incorrect ergonomic postures, may result in complications such as ocular fatigue, repetitive strain injuries, and diabetic retinopathy. Diabetes-related retinal damage (DR) is a condition that affects the eyes specifically and has the potential to impair vision. Stress is a psychological factor that can influence the onset and progression of diabetes and its complications. Touch screen data refers to the information collected from the interactions of users with touch-enabled devices, such as smartphones or tablets. Touch screen data can provide insights into the user’s behavior, mood, cognition, and health status. One of the new global concerns to public health is diabetes. Depending on the ophthalmologist's experience, the diagnostic process can be difficult or time-consuming, especially in environments with limited resources. Automated techniques are currently used to classify cases of Diabetes Retinopathy (DR). Tens of millions of individuals suffer from depression each year, yet only a small percentage of them get timely treatment. Therefore, it is essential to automatically identify human tension and relaxation using social media promptly. Early detection and effective management of stress are crucial in preventing its escalation into a severe condition. This paper presents a method to predict expressions of stress and Diabetic retinopathy by using ML techniques LR (logistic regression), RF ( random forest), and XG boost methods to detect the correlation between stress and diabetes while using touch screen.