Skip to main content
Built by KoderTroop · Computer Vision

EyeSight AI: Retinal Image Analysis Software

Computer vision that helps eye-care hospitals review retinal images, with the AI models and the application around them built by KoderTroop.

EyeSight AI screen
Computer Vision for Eye-Care Teams

Computer Vision for Eye-Care Teams

EyeSight AI analyzes retinal images to support the diagnosis of eye diseases in eye-care hospitals. KoderTroop built the software and the AI models. The models run inside an application designed around how a hospital captures, reviews and records images, so their output reaches clinicians where they already work. Every result informs a clinician's assessment, and the decision remains theirs.

How EyeSight AI Works

The screens your team works in each day, and what each one is for.

PT-10482 · right eye
Study ST-2291 · captured 09:12 · Camera 2
ODOSAR
Findings overlay
R1R2R3
OD · macula-centred · 45°3888 × 2592Image 1 of 2
PT-10482Analysed
Study ST-2291 · both eyes · waiting for review
Image qualityGood, gradable
Focus, field and exposure checked before analysis
Findings for review3 regions
Model output for a clinician to review. It is not a diagnosis.
R1
Small bright spots
Superior arcade
0.4 mm²
R2
Dark spot
Temporal to macula
0.2 mm²
R3
Vessel calibre change
Inferior arcade
1.1 mm
Send to reviewmodel v2.4.1 · analysed 09:13
Today's worklist
Eye clinic · capture, analysis, review and record
Study or patient IDAR
Captured
3
From camera rooms
Quality check
1
1 needs recapture
Analysed
4
Waiting for review
In review
2
With clinicians
Signed off
7
Today
Recorded
7
In patient record
StudiesAllMineNeeds recapture
StudyPatientEyesCapturedStageClinicianWaiting
ST-2292PT-10490OS09:18 · Camera 1CapturedUnassigned2 min
ST-2291PT-10482Both09:12 · Camera 2AnalysedARA. Rao8 min
ST-2290PT-10477OD09:05 · Camera 1In reviewMIM. Iyer15 min
ST-2289PT-10469Both08:58 · Camera 2RecaptureUnassigned22 min
ST-2287PT-10455Both08:40 · Camera 1AnalysedARA. Rao40 min
ST-2286PT-10448OS08:31 · Camera 2In reviewSDS. Das49 min
ST-2285PT-10441Both08:22 · Camera 1Signed offMIM. Iyer—
ST-2284PT-10436Both08:10 · Camera 2RecordedARA. Rao—
ST-2283PT-10430OD08:02 · Camera 1RecordedSDS. Das—
Review · PT-10482
Study ST-2291 · assigned to A. Rao
Waiting for sign-offAR
Both eyes · 14 Nov 2026Overlay shows regions for review
R1R2R3
Right eye (OD) · 3 regions
Left eye (OS) · no regions
Previous visitsCompare side by side
OD · 12 May 2026
OS · 12 May 2026
OD · 10 Nov 2025
OS · 10 Nov 2025
Findings for reviewModel output
Confirm or override each region. The clinician’s assessment is the record.
R1Small bright spotsConfirmed
Superior arcade, right eye
R2Dark spotOverridden
Temporal to macula · reason: camera artefact
R3Vessel calibre changeTo decide
Inferior arcade, right eye
ConfirmOverride
Clinician assessment
R1 confirmed. R2 is a camera artefact. Repeat imaging at the next routine visit and compare with today.
1 region still to decide before sign-off
ARA. RaoSign off
Pipeline
Image pipeline, model and application · last hour
1 hourToday7 daysKS
From camera to recordAll stages running
Intake
18 images0.3 s
Quality check
18 checked0.6 s
Pre-process
17 images0.4 s
Retinal model
17 analysed1.8 s
Review queue
6 waiting—
Report
7 recorded0.5 s
Recent studies
StudyImagesQualityModelTotalNow
ST-22921Passed—0.7 sAnalysing
ST-22912Passed1.8 s3.1 sReview queue
ST-22901Passed1.7 s2.9 sReview queue
ST-22892Recapture—0.9 sBack to camera room
ST-22882Passed1.9 s3.2 sReview queue
ST-22872Passed1.8 s3.0 sReview queue
ST-22861Passed1.6 s2.8 sReview queue
ST-22852Passed1.8 s3.1 sRecorded
Release 2026.11Deployed 9 Nov
Image pipelinev3.8.0
Retinal modelv2.4.1
Clinician appv1.12.0
Report servicev1.6.2
Built, tested and released together as one system.
Seconds per studyAverage 3.0 s
070809101112
Reports · PT-10482
Signed reviews are exported to the patient record
AR
Visits
14 Nov 2026Ready to export
ST-2291 · signed by A. Rao
12 May 2026Recorded
ST-1876 · signed by M. Iyer
10 Nov 2025Recorded
ST-1402 · signed by A. Rao
6 May 2025Recorded
ST-0988 · signed by S. Das
4 Nov 2024Recorded
ST-0571 · signed by M. Iyer
Export
PDF report
Images with overlay
Patient record systemConnected
Report previewPDFSend to record
Retinal imaging reportEyeSight AI
Patient
PT-10482
Study
ST-2291
Captured
14 Nov 2026, 09:12
Camera
Camera 2 · 45°
Reviewed by
A. Rao
Signed
14 Nov 2026, 11:06
Right eye (OD)
Left eye (OS)
Clinician assessment

R1 confirmed. R2 is a camera artefact. Repeat imaging at the next routine visit and compare with today.

Regions reviewed
  • R1 · small bright spots, superior arcade · confirmed
  • R2 · dark spot, temporal to macula · overridden
  • R3 · vessel calibre change, inferior arcade · confirmed

Model findings were reviewed by the signing clinician. This report records the clinician’s assessment.

Computer-vision models analyze each retinal image and highlight findings for a clinician to review.

Pricing

Quoted Per Deployment

EyeSight AI is priced to the scope of each deployment: sites, devices, integrations and the support you need.

Quote for EyeSight AI

Tell us how many sites and devices you operate and which systems it must connect to, and we will return a scoped quote.

Request a Quote

Common Questions