作者: Jun Cheng , Dacheng Tao , Jiang Liu , Damon Wing Kee Wong , Beng Hai Lee
DOI: 10.1007/978-3-642-23626-6_12
关键词: Cornea 、 Artificial intelligence 、 Angle-closure glaucoma 、 Optic nerve 、 Glaucoma 、 Trabecular meshwork 、 Open angle glaucoma 、 Computer vision 、 Feature (computer vision) 、 Iris surface 、 Iris (anatomy) 、 Iridocorneal angle 、 Optometry 、 Computer science
摘要: Glaucoma is an optic nerve disease resulting in loss of vision. There are two common types glaucoma: open angle glaucoma and closure glaucoma. type classification important diagnosis. Ophthalmologists examine the iridocorneal between iris cornea to determine type. However, manual classification/grading images subjective time consuming. To save workload facilitate large-scale clinical use, it essential automatically. In this paper, we propose use focal biologically inspired feature for classification. The surface located region. association grades built. experimental results show that proposed method can correctly classify 85.2% from 84.3% accuracy could be improved close 90% with more included training. effective automatic It used reduce ophthalmologists diagnosis cost.