SELECTION OF THE ARTIFICIAL INTELLIGENCE COMPONENT FOR CONSULTATIVE AND DIAGNOSTIC INFORMATION TECHNOLOGY FOR GLAUCOMA DIAGNOSIS

Abstract

The most important areas of application of consultative and diagnostic systems are urgent and life-threatening conditions characterized by a lack of time, limited opportunities for examination and consultations, and often little clinical symptoms with a high level of threat to the patient's life and the rapid pace of development of the process. The experience of using consultative and diagnostic systems proves a significant improvement in the quality of diagnostics, which not only reduces unjustified losses, but also allows more effective use of aid resources, regulates the volume of necessary research, and finally, increases the professional level of doctors for whom such a system serves at the same time and educational. Consultative diagnostic systems and technologies are currently rarely and insufficiently used in ophthalmology, although the field of ophthalmology in general and glaucoma diagnosis in particular are in great need of them.

Currently, the problem of using artificial intelligence for the problem of glaucoma analysis is faced with the fact that neural networks themselves and the methods of their use are not made suitable for mass use, with the complexity of development for certain models, with the inaccessibility for mass use, and the difficulty of collecting data for training neural models due to “confidentiality" of data. There is also the issue of cost and diagnostic availability — the availability of a trained professional, the means to collect data, and the time it takes for a patient to receive a diagnosis.

The author's further research will be aimed at creating the neural network itself for the diagnosis of glaucoma with different approaches from the available data types for each individual case, as well as creating programs and instructions for deploying such a neural network in places of use and using it with minimal requirements and resource needs. Compared to other similar products, this will be such an introduction of artificial intelligence that will allow to incorporate all the available experience into a small number of lines of code and will have pros in low budget and mass use.