Detection of Referable Diabetic Retinopathy using Machine Learning on Routine Clinical Data - Scorecard - MDSpire

Detection of Referable Diabetic Retinopathy using Machine Learning on Routine Clinical Data

  • By

  • Jeon, Young Joon

  • Song, Jae Shin

  • Borghare, Shubham

  • Lee, Youngju

  • Choi, Young Wook

  • Song, Junghan

  • Lim, Soo

  • Woo, Se Joon

  • April 13, 2026

  • 0 min

Share

Clinical Scorecard: Identifying Referable Diabetic Retinopathy through Machine Learning Analysis of Standard Clinical Data

At a Glance

CategoryDetail
ConditionReferable Diabetic Retinopathy (RDR)
Key MechanismsMachine learning algorithm utilizing clinical and laboratory parameters
Target PopulationAdults diagnosed with diabetes
Care SettingTertiary institution

Key Highlights

  • Machine learning model predicts RDR without ophthalmic imaging
  • Random forest model achieved AUROC of 0.932 in validation cohort
  • 15 significant predictors identified, with age being the most critical
  • Potential for early identification of RDR in resource-limited settings
  • Facilitates timely referrals and integration into clinical decision support

Guideline-Based Recommendations

Diagnosis

  • Utilize machine learning models for predicting RDR based on clinical data

Management

  • Implement early referral protocols for patients identified at risk of RDR

Monitoring & Follow-up

  • Regular assessment of identified predictors such as glycemic control and blood pressure

Risks

  • Inadequate monitoring may lead to vision impairment due to undiagnosed RDR

Patient & Prescribing Data

Adults with diabetes undergoing routine assessments

Focus on managing glycemic levels and cardiovascular risk factors

Clinical Best Practices

  • Incorporate machine learning tools into routine diabetic care
  • Utilize demographic and clinical data for risk stratification
  • Ensure regular follow-up for patients identified as high risk for RDR

Related Resources & Content

Original Source(s)

Detection of Referable Diabetic Retinopathy using Machine Learning on Routine Clinical Data

  • by Jeon, Young Joon , Song, Jae Shin , Borghare, Shubham , Lee, Youngju , Choi, Young Wook , Song, Junghan , Lim, Soo , Woo, Se Joon

  • April 13, 2026

Related Content

Diabetes Mellitus Coding in Retina

Age-related macular degeneration (AMD) is among the most common retinal diagnoses. Although the clinical features of the disease are well known, ICD-10 coding can be confusing if details such as stage, activity, and laterality are not clearly documented. Incomplete documentation may prompt questions about the medical necessity of diagnostic tests or intravitreal injections. This Q&A addresses common sources of confusion and offers practical guidance to support clear, accurate documentation.

Diabetes Risk Scores: Beyond Glucose

Multi-ancestry polygenic models identify high-risk patients traditional screening may miss.

Could Wegovy Raise Optic Nerve Risk?

A pharmacovigilance analysis of FDA adverse event reports finds the strongest ischemic optic neuropathy reporting signal with high-dose semaglutide formulations, with higher adjusted odds reported in men.