Artificial Intelligence System for Detecting Multiple Diseases Through Retinal Imaging - Scorecard - MDSpire

Artificial Intelligence System for Detecting Multiple Diseases Through Retinal Imaging

  • By

  • Xiayin Zhang

  • Qinyi Li

  • Yinhao Liang

  • Chunran Lai

  • Jiahui Cao

  • Yangqin Feng

  • Wenyi Hu

  • Hongyang Jiang

  • Chunxin Liu

  • Feng Zhang

  • Shan Wang

  • Ying Fang

  • Cuomu Duojie

  • Lumei Hu

  • Fan Xu

  • Kaiyi Chi

  • Miao Lin

  • Li Li

  • Yih Chung Tham

  • Yukun Zhou

  • Carol Y. Cheung

  • Xiaohong Yang

  • Bin Sheng

  • Zhuoting Zhu

  • Ching-Yu Cheng

  • Wing W. Y. Ng

  • Honghua Yu

  • April 28, 2026

  • 0 min

Share

Clinical Scorecard: Artificial Intelligence System for Detecting Multiple Diseases Through Retinal Imaging

At a Glance

CategoryDetail
ConditionEndocrine and metabolic diseases
Key MechanismsUtilizes retinal imaging and AI for early detection and monitoring
Target PopulationDiverse populations including multi-ethnic groups in resource-limited and high-resource settings
Care SettingPrimary care and clinical workflows

Key Highlights

  • Reti-Pioneer framework integrates multimodal datasets for disease prediction
  • Achieved AUROC values indicating strong performance in detecting T2DM, osteoporosis, and hypertension
  • Demonstrated generalizability across diverse populations and resource settings
  • Noninvasive approach potentially mitigates healthcare inequities
  • AI-driven model enhances diagnostic performance in collaboration with ophthalmologists

Guideline-Based Recommendations

Diagnosis

  • Utilize retinal imaging for early detection of endocrine and metabolic diseases
  • Incorporate AI models for improved diagnostic accuracy

Management

  • Implement Reti-Pioneer in routine primary care for disease screening
  • Focus on multimorbidity risk assessment in diverse populations

Monitoring & Follow-up

  • Regularly assess retinal changes as indicators of systemic health
  • Utilize AI for continuous monitoring and risk stratification

Risks

  • Ensure high-quality imaging to avoid misdiagnosis
  • Address potential biases in AI models due to underrepresentation of certain populations

Patient & Prescribing Data

Individuals at risk for endocrine and metabolic diseases

AI-driven insights can guide personalized treatment plans based on retinal assessments

Clinical Best Practices

  • Adopt a unified framework for simultaneous identification of multiple conditions
  • Leverage AI to enhance traditional screening methods
  • Conduct regular training and validation of AI models across diverse datasets

References

Original Source(s)

Artificial Intelligence System for Detecting Multiple Diseases Through Retinal Imaging

  • by Xiayin Zhang, Qinyi Li, Yinhao Liang, Chunran Lai, Jiahui Cao, Yangqin Feng, Wenyi Hu, Hongyang Jiang, Chunxin Liu, Feng Zhang, Shan Wang, Ying Fang, Cuomu Duojie, Lumei Hu, Fan Xu, Kaiyi Chi, Miao Lin, Li Li, Yih Chung Tham, Yukun Zhou, Carol Y. Cheung, Xiaohong Yang, Bin Sheng, Zhuoting Zhu, Ching-Yu Cheng, Wing W. Y. Ng, Honghua Yu

  • April 28, 2026

Related Content

Keeping an Eye on Diabetes

Progress, pitfalls and the future of the UK’s Diabetic Eye Screening Program

A clinically interpretable machine learning model for early detection of diabetic retinopathy in multiple community health centers

Metformin’s Potential Role in Posterior Capsule Opacification

Common diabetes therapy reaches intraocular lens tissue and may help limit PCO, says newly published data

  • March 30, 2026

  • 3 min