分頁p2智視康_底圖1.jpg
智醫官網_分頁_智視康綠色logo.png

 

VeriSee DR

AI-assisted solutions for Diabetic Retinopathy (DR) identification

智醫官網_icon1.png
No ophthalmologist required
Can be expanded for use in community care
智醫官網_icon2_new.png
Quick screening for
high-risk cases
Helps ensure timely treatment
智醫官網_icon3_new.png
Integration with
medical record systems
Cross-disciplinary use

Improving the diagnosis and tracking process for DR

分頁p2智視康_底圖3.jpg
VeriSee DR_Main UI_20200427.jpg

VeriSee DR is an AI diagnostic aid that has been trained using DR diagnosis experience from multiple ophthalmologists and uses AI deep learning techniques to produce diagnosis results similar to that of professional physicians. Acer Healthcare Inc. uses hardware integration to construct an edge computing device that can be used without a network connection.

AI-based Diagnosis Software for Diabetic Retinopathy

by Dr. Hsieh, Yi-Ting

 

Diabetic Retinopathy, DR

輪播1.png

Normal Eye

Eye with Retinopathy

Healthy blood vessel

Tiny blood vessels
leak fluid into the retina

Diabetic Retinopathy (DR) can cause blindness

If diabetic patients fail to control their blood glucose levels and seek continuous medical attention, as time progresses the likelihood of developing DR increases. 20–30% of patients with Diabetic Retinopathy are at risk of going blind, with an average age range of 20–65.

 

With instant photographic analysis, testing and diagnosis can be conducted in one session

Capture an image and press analyze to view the results.
Install VeriSee on a computer with an ophthalmoscope connection and select a folder location for fundus image exports. Each time a fundus image is added to the folder, the image is automatically uploaded to VeriSee for quality verification. After image quality has been verified by the user, the image can be analyzed.

105273799_m.jpg

Batch analysis mode can be used to efficiently handle a large numbers of cases

Up to 1000 fundus images can be uploaded to VeriSee at once. Analysis files are saved in CSV format for convenient connection to hospital systems and for use in relevant research.

Exporting the fundus image and analysis report helps ensure smoother doctor-patient communication

Following analysis, reports can be edited and exported in PDF format.

資訊系統整合EN.png
分頁p2智視康_底圖3.jpg

VeriSee DR AI technology

After each image is uploaded

3 seconds

*

Assess whether there is a risk of diabetic retinopathy and need referral.

The clinical trial results reached 

95%

Sensitivity

90%

Specificity

2018 12th Asia-Pacific Vitreo-retina Society (APVRS) Congress,

and 2019 Asia-Pacific Academy of Ophthalmology(APAO) Annual Congress

Journal of the Formosan Medical Association (2020). Application of deep learning image

assessment software VeriSee for diabetic retinopathy screening. (IF 2.844)

https://www.sciencedirect.com/science/article/pii/S0929664620301182

Life (MDPI, 2021) . The Clinical Influence after Implementation of Convolutional Neural Network-Based Software for Diabetic Retinopathy Detection in the Primary Care Setting. (IF 2.991)
https://www.mdpi.com/2075-1729/11/3/200

ClinicalTrials.gov
https://clinicaltrials.gov/ct2/show/NCT04160988?term=verisee+dr&draw=2&rank=1

* in a CPU Core i7 and GPU environment