Smartphone-based AI-supported telemedical diabetic retinopathy (DR) screening
Partner organisation in partner country: Sankara Eye Foundation India (India), University of Calabar Teaching Hospital (Nigeria), Organization for Rural Community Development (Bangladesh), University of Cape Coast Ghana Department of Optometry & Vision Science (Ghana)
Situation:
Diabetic retinopathy (DR) is the most common cause of blindness in working-age adults worldwide. However, 80% of diabetics live in low- and middle-income countries, where DR screening is largely unavailable. The organisation has established a low-cost, smartphone-based telemedical DR screenings in India, Bangladesh, Nigeria, and Ghana at different project stages.
Objectives:
Demonstrate the scalability, cost-effectiveness, and sustainability of smartphone-based AI-assisted DR screening in the regions of the project partners.
- Improved DR screening coverage by 50%,
- Published cost-effectiveness analysis
- Established standardized processes and multipliers, as well as country-specific cost compensation
- Project expansion in all partner regions and implementation of 572 DR screening camps
- Establishment of AI-assisted data collection using standardized surveys (regarding screening coverage, referral rates, cost efficiency)
- Training of 24 multipliers, SOP generation and implementation for all project steps, gradual integration of cost compensation
The establishment of regular local DR-screenings and the training of multipliers, as well as the transfer of knowledge for independent training implementation, will enable the project to be continued in a sustainable manner over the long term. The publication of the cost-efficiency analysis is a prerequisite for long-term implementation in local health systems, both in the partner countries and in other countries of the Global South.
Smartphone-based fundus imaging is a low-cost method for digitally imaging the retina. Due to the low purchase cost of the devices, their high availability, and their ease of use even by non-medical personnel, it can be implemented relatively easily even in resource-poor contexts (both in terms of financial resources and medical personnel). Studies have repeatedly proven the medical effectiveness and cost-efficiency of telemedical DR screening, but the high cost of fundus photography devices is a hurdle that has so far prevented the widespread implementation of DR screening in the Global South. Smartphone-based fundus imaging offers a possible solution to this problem. It enables ophthalmologists to use their limited resources more efficiently, as they only have to assess the transmitted retinal images and no longer have to perform the actual screenings themselves. In previous projects in India, Bangladesh, Nigeria, and Ghana, we have already established low-cost telemedical smartphone-based DR screenings in various project stages. In addition, we are developing an AI for the automatic analysis of smartphone-based retinal video recordings to support image evaluation. As part of this project, the scalability, cost-effectiveness, and sustainability of the smartphone-based and AI-supported approach is now being investigated in the countries mentioned.
Further information here.
