Artificial Intelligence Based Glaucoma Detection

Abstract

Artificial Intelligence has shown significant enhancement recently, in many fields, especially in pattern recognition in computer vision and hence in medical diagnosis, wherein highly accurate automated diseases diagnosis has now become feasible. Artificial Intelligence methods, especially based on deep learning are very powerful and have got huge training and learning potential. These methods are growing from strength to strength and far exceeding the performance of traditional methods. The deep learning has proved to be highly efficient and accurate especially when balanced and large training data is available. This chapter provides an overview of various Artificial Intelligence based techniques for detection of the second-largest disease which causes blindness, that is, glaucoma. The chapter begins with the introduction of glaucoma and the various important parameters required for its accurate diagnosis. The later sections introduce various traditional Machine learning based techniques, followed by Deep learning based techniques including the Transfer learning and Ensemble learning based techniques in use for automated glaucoma detection, along with their advantages and drawbacks. This chapter also exposes readers to original and genuine data of perimetry and OCT tests.