A Digital Twin Framework for Applications in Ophthalmology

A Digital Twin represents a physical entity through simulation of it's data, behavior, and communication within the physical environment. Digital Twin constitute a digital replica of the corresponding object or process, being able to provide their real-time assessment and monitoring. Digital Twin founds many industrial applications, but they are also increasingly used in healthcare, where they already have a relevant utility and impact. In healthcare systems, Digital Twins is considered as an important progress factor which revolutionizes the patient care providing mode. In this paper, a Digital Twin framework for healthcare application, focusing on ophthalmology is presented. Given the high level of complexity in creating a Digital Twin associated with the eye due to the optical properties of the visual system, a neural network has been trained to identify glaucoma based on a dataset. Glaucoma detection was performed by processing images and extracting the cup/disc ratio, which, at the time of pathology detection, has values ranging from 0.6-0.8. A prediction was realized using linear regression to assess the progression of the patient's pathology over the next 5 years.