v1.0.1

80% of all blindness can be prevented¹

And the first step is early and accessible screening. RetinaLyze® is a screening software, which helps eye care professionals to perform screenings for pathological changes in the retina and optic nerve head that could potentially indicate early signs of eye diseases in a safe and efficient manner. It can be used as a clinical decision support system (CDSS) or as a screening tool in conjunction with the telemedicine module.

Open

The RetinaLyze® system combines the speed and efficiency of Artificial Intelligence with the guidance and supervision of telemedicine to dramatically lower the cost of eye screening and makes it possible to offer treatment to patients around the world at manageable costs. The RetinaLyze® system also enables retail eye care professionals such, as optometrists, to perform safe and easy eye screenings, thereby increasing the availability of eye-screenings for the general public.

The RetinaLyze screening process

A RetinaLyze screening is performed by capturing a fundus image or OCT-scan and uploading it to the RetinaLyze service. You can run an analysis for investigation of potential pathological changes like minor black and white lesions on the retina, and colour variations in the optic nerve head. The identified potential pathological changes could indicate early signs of eye diseases like Diabetic Retinopathy (DR), Age-related Macular Degeneration (AMD), and Glaucoma and over 70 signs of eye diseases on OCT-scans.

The results are shown to the operator within 30 seconds. The entire process can be initiated using a web-browser, internet connected software (such as EHR/EMR/Patient journal systems) or an internet connected fundus camera.

The result is given in four easy to understand colour codes: Red, Yellow or Green and Grey. Red, Yellow and Green indicates the severity of the found changes and Grey indicates that the image was not of acceptable image quality to perform the analysis. It will also be possible to see an overlay of the detected retinal changes on the fundus image.

When screening the general public, there will be no detected changes in 85-90% of the cases and the operator can store the image and ask the patient to return next year for a new test. 

In the event of detected retinal changes the image must be sent to a specialist, who will provide the operator with a manual assessment. In some cases, the patient will be advised to contact a local ophthalmologist for further examination. This ophthalmologist then performs the final diagnosis and treatment.

Types of automated analysis

The RetinaLyze System consists of vector- and colorimetry-based algorithms. The patented invention relates to a method for assessing the presence or absence of retinal lesion(s) and damage to the optical nerve head in a fundus image. 

Retinalyze can run multiple types of analyses identified in the system by the abbreviations DR, AMD, Glaucoma and OCT.

RetinaLyze DR is made to focus on investigation of potential pathological changes like minor black lesions on the retina, which could indicate early signs of eye disease like Diabetic Retinopathy (DR). RetinaLyze DR detects typical lesions of diabetic retinopathy like micro-aneurysms and minor hemorrhages.
Read more here: RetinaLyze DR

RetinaLyze AMD is made to investigate the potential pathological changes like minor white lesions on the retina, which could indicate early signs of eye disease like of Age-related Macular Degeneration (Dry non exudative AMD).
Read more here: RetinaLyze AMD

RetinaLyze Glaucoma analyses a fundus image to detect signs of glaucoma. The algorithm works by measuring the hemoglobin (Hb) levels in the optic nerve head (ONH) by means of color fundus images (colorimetry).

Read more here: RetinaLyze Glaucoma

RetinaLyze OCT analyses OCT scans and detects over 70 signs of pathology in 45 seconds. Read more here: RetinaLyze OCT

SOURCES

1. https://www.emro.who.int/control-and-preventions-of-blindness-and-deafness/announcements/action-plan-prevention-avoidable-blindness-visual-impairment-2014-2019.html