Our offer

SCREENING TESTS

CE agreed Medical Device, our solution, in its current versions, detects:
Diabetic Retinopathy and Diabetic Macular Edema; Age related Macular Degeneration and Age related Maculopathy; and Glaucoma

Diabetic Retinopathy (DR)

Diabetic Retinopathy and Diabetic Macular Edema together affect 20 to 50% of diabetics, for a total of nearly 150 million people worldwide.

SCREENING TESTS

CE agreed Medical Device, our solution, in its current versions, detects:
Diabetic Retinopathy and Diabetic Macular Edema; Age related Macular Degeneration and Age related Maculopathy; and Glaucoma

Glaucoma

Glaucoma affects about 80 million persons worldwide, and is the first cause of cecity in people over 60.

Age-Related Macular Degeneration (AMD)

Age related Macular Degeneration affects 200 million persons worldwide. Its early form is Age related Maculopathy.

Retina image quality analysis

Fundus photography quality check

  • OphtAI checks in real time for submitted photos’ quality to verify it can be interpreted.
  • Hence, the users know while screened persons are still present and available if they have to capture new images.
  • This function allows using any fundus camera on the market, as long as its photos’ quality is sufficient.

Lesion mapping

  • Our innovative technology, designed by our scientists, allows lesion mapping.
  • This is the so-called heatmaps.
  • They display the importance of each pixel in OphtAI’s decision.
  • This is graphically represented as a proportionally intense luminosity.

Glaucoma Example – Slide to discover the “gradual borders” type heatmap

Performances

Diabetic Retinopathy

Glaucoma

AMD

Sensibility

99%

96.5%

92.5%

Specificity

87%

95.4%

87%

Response time

<3s

<3s

<3s

Classification and Proportion of Similar Cases

  • Depending on the proportion of images receiving a given score for each ground truth (usually, healthy or not for a given disease) on a database whose cases are balanced as a function of the disease’s prevalence ratio, we process the confidence level for the image score.
  • A confidence level equates Positive predictive value for a positive diagnosis and a Negative Predictive Value for a negative diagnosis.
  • So, if the patient is to be refered to an ophthalmologist (CNN output score >= threshold, confidence matches the probability for the patient actually being affected by the disease; this probability is low close to the threshold and reaches 100% when the score increases.

Grading

  • OphtAI is able to grade DR’s severity stage
    >>This function is very important as it allows knowing the appropriate course of action and the emergency level for the patient to meet an ophthalmologist.

PDF Report

  • Our report is a Medical Device certified PDF allowing for diagnosis,
  • available from any internet browser without external application,
  • it is customizable and allows on demand anonymization, per disease results and confidence levels integration, and lesion mapping and importance integration.
Our eye disease detection service,
100% cloud, 100% autonomous, easy to use and immediate.

 

Notice

A major disorder of the transparent media, such as cataract, does not allow reliable analysis of the retina by retinophotography: in this case, use of the device should be contraindicated.

Have a question ? 

Share This