Red Reflex Africa Smartphone-based childhood eye screening
For screening and educational use only. This application does not provide a medical diagnosis.

Red Reflex Africa

Save the life of your child — with your smartphone

A screening, triage and referral support tool for health workers and parents. It photographs a child's red reflex with the phone camera and flash and helps decide who needs an eye examination.

Initiated by Dr. med. Harald C. Gäckle · Eagle Eye Laser Centre, Nairobi

© 2026 Dr. Harald C. Gäckle · Eagle Eye Clinic Nairobi · All rights reserved

Refer without delay if parents report:

    If parents report a white pupil, refer the child even if the current photograph appears normal.
    Demonstration mode
    Test the workflow with simulated example images. All results are labelled as demonstrations.
    Step 1 / 6

    Child information

    Screening ID
    Use anonymous screening ID only
    No name or contact details are stored.

    Medical history

    Step 2 / 6

    Warning signs

    Step 3 / 6

    Prepare the child and the room

    Step 4 / 6

    Capture flash photographs

    Take at least 3 images (centre, slight right gaze, slight left gaze). Multiple images reduce the risk of angle-dependent photographic reflections.

    Position at 1–1.5 m distance. Both pupils should fill a good portion of the screen. Use zoom to magnify.

    No continuous light — a short flash fires only when you press the shutter, so the pupil stays wide.

    When on: a search beep sounds, then a steady tone and automatic shutter once both pupils are inside the circles.

    Images captured: 0 / 5

    Step 5 / 6

    Image quality assessment

    A poor or non-assessable image never receives a reassuring green result. Unable to assess safely → repeat the photograph or refer if clinical concern remains.
    Step 6 / 6

    Red reflex analysis

    Manual pupil correction: drag the R and L circles exactly over the pupils. Pinch or use the slider to resize.

    Screening History

    Referral Centres

    Education

    📷 Reference images

    Example photographs for training and demonstration. Tap an image to run the full analysis on it — useful for showing colleagues how the app behaves on a known picture.

    Built-in examples

    Your own images

    Stored only on this device, in this browser. They are never uploaded anywhere. Clearing the browser data deletes them.

    ⚡ Quick screen all images

    Runs the full analysis on every image above — built-in examples and your own — and lists the result for each. Tap a row to open that image in the normal analysis. For teaching and for checking the app against a known set; not a substitute for examining a child.

    In the picker you can select many photos at once — open an album and tap Select. A browser cannot read the photo library on its own, so this selection step is required; everything after it runs automatically.

    Settings

    General

    Screening sensitivity

    Accessibility & display

    Large text
    High contrast
    Dark mode
    Vibration for urgent results
    Spoken instructions (where supported)

    Data protection

    Data is stored only on this device. Nothing is uploaded or synchronised automatically. Export and sharing are always manual actions.

    Anonymous mode by default
    Delete images after referral
    Retain only eye crops

    Encrypted local export: placeholder — planned for a future version.

    Validation dashboard (administrator prototype)

    For future validation studies: enter the ophthalmologist result for referred children in the record's follow-up section. Metrics below compare app result (abnormal = category ≥ 2) with clinical outcome.

    The intended development priority is minimising false-negative leukocoria results.

    Application data

    About & Safety

    Safety rules

      About this application

      Red Reflex Africa is a prototype for research, education and clinical validation. It is not a certified medical device and must not be used for autonomous diagnosis. The demonstration analysis is a transparent rule-based algorithm; all thresholds are placeholders that require prospective clinical validation. A future version is prepared to integrate a real AI model (TensorFlow.js or external API).

      Initiated by Dr. med. Harald C. Gäckle, ophthalmologist and refractive surgeon, Eagle Eye Laser Centre, Lavington, Nairobi, Kenya.

      Translations must be reviewed by native-speaking medical professionals before clinical deployment.

      Version 0.9 prototype · Runs offline after first load. Data stays on this device.