Diagnostic Efficacy of CNN in Differentiation of Visual Field

Diagnostic Efficacy of Convolutional Neural Network Based Algorithm in Differentiation of Glaucomatous Visual Field From Non-glaucomatous Visual Field

Glaucoma is currently the leading cause of irreversible blindness in the world. The multi-center study is designed to evaluate the efficacy of the convolutional neural network based algorithm in differentiation of glaucomatous from non-glaucomatous visual field, and to assess its utility in the real world.

Glaucoma is the world's leading cause of irreversible blind, characterized by progressive retinal nerve fiber layer thinning and visual field defects. Visual field test is one of the gold standards for diagnosis and evaluation of progression of glaucoma. However, there is no universally accepted standard for the interpretation of visual field results, which is subjective and requires a large amount of experience. At present, artificial intelligence has achieved the accuracy comparable to human physicians in the interpretation of medical imaging of many different diseases. Previously, we have trained a deep convolutional neural network to read the visual field reports, which has even higher diagnostic efficacy than ophthalmologists. The current multi-center study is designed to evaluate the efficacy of the convolutional neural network based algorithm in differentiation of glaucomatous from non-glaucomatous visual field, compare its performance with ophthalmologists and to assess its utility in the real world.

Inclusion Criteria:

1. Age≥18;
2. Informed consent obtained;
3. Diagnosed with specific ocular diseases;
4. Able to perform visual field test

Exclusion Criteria:

Incomplete clinical data to support diagnosis