Starbugs: Diabetic Retinopathy Detection (Phase I)
PROBLEM
Indonesia currently has the world’s fourth largest population, and ranks seventh when it comes to people diagnosed with diabetes mellitus . Among 35% of adult diabetics are affected by diabetic retinopathy (DR), a side effect of the disease. Simultaneously, the lack of check-up facilities and early warning systems for diabetics, compounded with DR’s asymptomatic nature which often leads to infrequent follow-ups, have commonly resulted in cases where the disease is no longer treatable, causing permanent blindness. This makes DR a leading cause of loss of vision in working age adults; on a more positive note, such blindness is almost completely preventable with appropriate screening and treatment. The challenge, in this regard, is to improve compliance among diabetic patients with diabetic retinopathy, as well as the screening system for diabetic patients.
The team worked with the organization Helen Keller International (HKI) Indonesia to assist them with compliance improvement, logistics and management related to DR treatment. SOLUTIONThe project team has successfully developed the basic interface of the app called “SnapGrade” as well as the basic detection algorithm that recognizes retinal features of a diabetic retinopathic eye. This helped HKI to save time during screening processes and alleviated the problem of manpower shortage in analyzing and grading retinal images for DR.
The team went to Indonesia to test their prototype and get a better understanding of HKI’s workflow. During the visit they learnt about the screening process at the primary clinics as well as the flow of patients and information to tertiary hospitals during the treatment process. The team visited clinics in Jakarta, Bandung and talked with doctors to understand their pain points, and receive feedback on the prototype. |