Abstract:
Ophthalmology is the branch of medical science that deals with treatment and surgery of the
disorders that occur inside the eye. These disorders can be in the form of Diabetic Retinopathy,
Glaucoma, AMD, Cataract etc and they constitute most of the cases of blindness which can be
avoidable (80%). The combined total of the total number of people who are visually impaired
people is 285 million and the number is expected to increase soon if no proper measures are taken.
In Pakistan alone we have 480 people who daily develop some sort of eye disease so a cloud-based
solution can work with 4G infrastructure in the country and provide access to any remote area.
Also, the other problem is that in countries like Pakistan, a huge majority of graduates of Medicine
are Females yet only 50% of them can practice after graduation due to social and cultural
limitations. So a remote cloud based solution can help them practice ophthalmology while sitting
at their homes. The diagnosis of ophthalmology disorders are done using two major techniques i.e
OCT Scans and Fundus Image Examination. Previously, the whole methodology of treating and
diagnosis of these disorders involved a cycle of multiple visits to the doctor back and forth. This
project involved a solution which automates the long diagnosis cycle for these disorders and is
beneficial for all stake holders – mainly focusing on a cloud based expandable system.
The multiple modules which will be explained in the next part of the report can be used to replace
the complete manual diagnosis by AI and Deep Learning based techniques. The doctors,
ophthalmologists, graders, and adjudicators can login into the cloud-based platform and grade
images from anywhere in the world. They are also able to provide their feedback which is used to
perform Incremental Learning. The platform also gives complete reporting options for patients
based on clinical findings.
