ISB

Artificial Intelligence-Enabled Referral Pathways for Early Diagnosis of Complications Arising from Diabetes: Application to Diabetic Retinopathy

Artificial Intelligence-Enabled Referral Pathways for Early Diagnosis of Complications Arising from Diabetes: Application to Diabetic Retinopathy

Year: March 2024

Collaborators: IDEA clinics, LV Prasad Eye Institute (LVPEI), Indian Council of Medical Research (ICMR)

Team: Sarang Deo, Swetlina Hota

Share:

Background

Diabetic Retinopathy (DR) is one of the leading causes of vision impairment among diabetic patients. A complication of diabetes, DR is preventable through timely screening and diagnosis. To combat the delay in diagnosis through traditional methods, our project aims to implement Artificial Intelligence (AI)-enabled systematic screening at diabetic clinics to enhance accuracy of diagnosis and streamline referral pathways to ophthalmology clinics for timely intervention and treatment.

About the Study 

AI-enabled screening systems will be installed in IDEA clinics where diabetic patients are monitored and treated. The study will assess whether these systems assist in earlier detection of visual impairment in patients in comparison to traditional tertiary care methods. Besides evaluating the effectiveness of AI enabled screening, its cost-effectiveness and its impact on follow-up adherence among patients will also be assessed.

Methodology

The study will use a prospective cohort design, wherein the diabetic patients recruited at the various diabetic clinics will be screened with AI-enabled cameras by trained non-ophthalmic technicians and nurses. Outcomes for the study cohort will then be compared to the standard care recipients at tertiary eye clinics. Key evaluation metrics will include:

  • Accuracy of DR diagnosis
  • Referral rates
  • Treatment uptake
  • Patient satisfaction
  • Cost effectiveness analysis to assess the economic viability of AI assisted screening programs

In nutshell, the study aims to establish the usefulness and economic viability of using AI technology in DR screening and diagnosis.


Intended Outcomes

By establishing the effectiveness of AI-enabled systematic screening for DR, the study will generate evidence to facilitate early detection to reduce the risk of vision loss among diabetics. Additionally, the study will generate insights to support policies and practices for improving DR management. Ethical considerations like informed consent and participant welfare, will be paramount to ensure sustainable healthcare advancement and responsible integration of technology into healthcare.