Examinations of the back of the eye can be vital to treat patients before vision loss or other symptoms when considering multiple eye conditions:
- Macular degeneration
- Glaucoma
- Diabetic Retinopathy
- Other vision loss
- Retinal vein occlusion
- Retinal detachment
- Retinal tears
By providing 50+ retinal cameras, Seva is helping to expand access to imaging and collaborating on managing these new patients through the referral, diagnosis and treatment journey.
We want to explore various patient journeys and will use an example of 3 friends who are aware of the risk for retinal conditions to illustrate the risk of missed imaging or narrowly-scoped referral algorithms.
To explore the patient experience, imagine 3 friends, all of them female in their 40s, who live in a remote village. Their names are Poppy, Iris and Lily.
- Awareness: The three women see an online announcement for an eye screening event and decide to go together. They know vision risks may increase as they grow older and want to learn more and get their eyes examined. Poppy had to work late and was not able to attend the screening as planned.
Poppy: We see the first divergence of paths as Poppy misses the retina imaging and has an unknown outcome.
- Retina image: There is a retina camera at the screening event as more portable, durable, affordable cameras are on the market.
- Algorithm: There is not an ophthalmologist at the screening event so the team decided to use automated image analysis, like artificial intelligence algorithms, to screen people. The algorithm been trained for diabetic retinopathy and properly refers Iris for treatment. However, it is not trained for macular degeneration and Lily does not understand that “no diabetic retinopathy” is not equivalent to “normal.”
Now the friends’ paths diverge.
Iris
IA. Treatment: Iris is referred to the city hospital for treatment. She makes the journey and is not able to afford anti-VEGF intravitreal injections but can get laser therapy to help her retinopathy. She is also advised on how to better control her diabetes.
IB. Follow-up: Iris returns to the city hospital for additional examination to confirm that her diabetic retinopathy has improved and is shown her retinal images over time to encourage her to continue managing her diabetes.
Lily
L. Missed Referral: Lily’s evaluation by the algorithm indicated she did not have diabetic retinopathy but she did not realize her retina image should be evaluated for other conditions.
Lily needs clear counseling and to understand that different condition could be found later that requires follow up. If she misunderstands or isn’t educated, she could believe that her eyes are normal and miss treatment and follow up.
Seva invites your feedback! Have you used an algorithm to assist with referral or diagnosis? What are your priorities in selecting an algorithm for future use?
Several companies have released algorithms for retinal imaging. Seva has not tested these algorithms but welcomes partner questions or feedback as we continue to explore this space.
| Algorithm | Conditions Referred | Cameras Supported |
| ARDA, Google | Referable DR or macular edema | Centervue DRS, Optovue iCam, Canon CR1/DGi/CR2, Topcon NW using 45° fields of view |
| Remidio Medios | 1. DR 2. Glaucoma | Remidio Fundus on Phone |
| IDxDR | Mild to moderate DR in adults with diabetes | Topcon NW400 |
| EyeNuk – EyeArt | DR | Canon CR-2 AF, Canon CR-2 Plus AF, and Topcon NW400 |
| AEYE | 1. DR 2. AYECS – comprehensive screening (Multiple conditions) | Optomed Aurora portable handheld device |
Please comment below with your comments or questions – we’d love to hear what’s important to you!
Leave a Reply