Live demo

Build a patient card, live

Add a fundus photo, an OCT scan and the patient's numbers: any of them, in any combination. Each part of the card fills in as soon as its input is analysed, and the phenotype appears once IOP and blood pressure are known.

Fundus AI trained on Chákṣu· openvalidated on PAPILA + HYGD (unseen cameras)· open
OCT cup depth image measurement, checked against the clinic's HRT onLSMU· real
Screening trained on PAPILA· openpercentiles against LSMU NTG cohort· real

Fundus photo

optional
Drop a fundus photo
or click to choose · JPEG or PNG

Open-data samples also fill in age, sex and IOP from PAPILA.

OCT scan through the disc

optional

One B-scan image, or a sweep video: then the deepest cup is found for you.

Drop a B-scan or a sweep video
JPEG, PNG or MP4

Synthetic scans. The sweep is a series of frames; the deepest cup is found for you.

Patient details and vascular tests

optional
Examples (not patients):
Flicker example:

IOP and blood pressure give the phenotype; the rest adds to it.

Patient card
Built live from whatever you add on the left: a photo, a scan, numbers, or any mix of them.

Structure · fundus photo

No fundus photo. Add one to outline the disc and cup.

Structure · OCT scan

No OCT scan. Add a B-scan or a sweep video through the disc to measure the cup.

Pressure and vascular function

Pressure
add IOP
Perfusion
add IOP and blood pressure
Pulsatility
add a Sonovum amplitude
Flicker response
add a flicker test result

Phenotype hypothesis

Add IOP and blood pressure to see what may drive this glaucoma.

Percentiles against the real LSMU NTG cohort. The flicker axis uses a flicker test result when one is given. Research prototype, not a diagnosis.