VeriSee DR

VeriSee DR is an AI-assisted diagnostic software for diabetic retinopathy (DR) that analyzes fundus camera images to identify high-risk patients requiring ophthalmologist referral. Developed by Acer Medical with core AI models trained on images labeled by retina specialists, the software achieved 93% screening accuracy in clinical validation.

Product Overview

As the first AI ophthalmology Software as a Medical Device (SaMD) approved by Taiwan FDA, VeriSee DR addresses the global challenge of limited ophthalmologist availability for diabetes patients requiring annual eye examinations. The software enables primary care facilities and diabetes clinics to conduct DR screening without specialized ophthalmology expertise, providing immediate referral recommendations comparable to professional ophthalmologist assessment.

The system processes fundus images in real-time, typically outputting recommendations in less than one minute. It operates without network connectivity, ensuring patient data security while enabling deployment in remote healthcare settings.

Key Features

For diabetic retinopathy screening, Acer Medical’s VeriSee DR identifies four lesion types in fundus images with 93% accuracy and outputs referral recommendations in under one minute.

• Lesion Detection: Identifies four major DR lesions—microaneurysms, hemorrhages, soft exudates, and hard exudates—with corresponding location mapping
• Batch Processing: Analyzes up to 1,000 images in a single session for high-volume screening programs
• Multi-Brand Compatibility: Supports fundus cameras from Topcon, Canon, Crystalvue, Nikon, and MiiS
• System Integration: Direct connectivity with hospital HIS and PACS systems
• Offline Operation: Runs on local AI PC without internet connection

Technical Specifications

VeriSee DR’s offline Intel Core Ultra platform processes up to 1,000 fundus images per session without network connectivity, supporting deployment in remote clinics.

Clinical Applications

VeriSee DR targets diabetes clinics and primary care facilities where diabetic patients receive regular metabolic management but may lack access to ophthalmology services. Healthcare professionals capture fundus images using compatible cameras, and the software immediately provides two-tier classification: “Regular tracking” or “Referral recommended.”

The lesion detection feature allows cross-referencing of referral recommendations, reducing both false positive and false negative results. This dual-output approach supports clinical decision-making while maintaining efficient patient throughput.

Regulatory Status

Market Deployment

VeriSee DR has achieved approximately 80% adoption rate among Taiwan medical centers, with implementation exceeding 100 healthcare institutions domestically. Taipei City Hospital deployed the system across six branches including Zhongxing, Heping, Renai, Zhongxiao, Yangming, and Linsen facilities. In Thailand, the software operates in over 30 hospitals, particularly serving diabetic patients in remote northeastern provinces and high-volume facilities in Bangkok.

Frequently Asked Questions

What is VeriSee DR’s screening accuracy?

VeriSee DR achieves 93% screening accuracy for diabetic retinopathy referral recommendations, based on results from a pivotal clinical trial approved by Taiwan FDA and published in the Journal of the Formosan Medical Association.

Which fundus cameras are compatible with VeriSee DR?

VeriSee DR supports multiple fundus camera brands including Topcon, Canon, Crystalvue, Nikon, and MiiS. Acer Medical also offers the ARC-1000 non-mydriatic fundus camera as part of an integrated screening solution.

Does VeriSee DR require internet connectivity?

No, VeriSee DR operates entirely on local hardware without network connection, ensuring patient data privacy and enabling deployment in areas without reliable internet access.