A mobile app for Glaucoma diagnosis and its possible clinical applications.

BACKGROUND

Nowadays, the latent power of technology, which can offer innovative resolutions to disease diagnosis, has awakened high-level anticipation in the community of patients as well as professionals. An easy-to-use mobile app is developed by us, which is purposefully intended for those patients with glaucoma.

METHODS

A mobile App has been invented for smartphones for the convenient use wherever and whenever. The corresponding experiments carried out by public retinal image database and real captured clinical data reveal the ideal classification accuracy of the App. Also, user feedback evaluation is also carried out in terms of performance test as well as and users' experience.

RESULTS

For clinical test using Yanbao App, we found 274 patients for the identification with 648 retinal images to be evaluated by glaucoma classification. Of the 243 glaucoma patients, 191 were screened out with an accuracy of 0.7860 (sensitivity); the number of non-glaucoma patients was 310 of 405, and the accuracy reached 0.7654 (specificity).` The total Accuracy amounted to 0.7731, and the result is close to the test performance obtained on public dataset ORIGA and DRISHTI-GS1.

CONCLUSIONS

Yanbao App can be applied as an innovative approach exploiting mobile technology to enhance the clinicians' efficiency and a balanced medical resources as well as a provided better tiered medical service system.