LungAI
Educational Research Demo

Early detection,
measured in seconds.

LungAI is an open-source CNN that screens lung CT slices for signs of cancer. Upload a scan and get a confidence score in under a second.

NOT A MEDICAL DEVICE · CONSULT A LICENSED CLINICIAN

CT-SLICE-001
224×224
CONFIDENCE 94.2%
● ANALYZED
80%
Stage 1 5-yr survival
<5%
Stage 4 5-yr survival
70%
Cases caught too late
1 in 5
Cancer deaths globally
PIPELINE

Three steps. One result.

Upload, analyze, review. The whole loop runs locally against your backend instance.

STEP 01

Upload

Drop in a CT slice (JPG or PNG). Files never leave memory.

STEP 02

Analyze

A CNN runs inference on the slice in under a second.

STEP 03

Review

See a binary classification with a confidence score and raw model output.

Sub-second inference

TensorFlow + FastAPI on the backend. No queues, no waiting.

Zero retention

Uploads are processed in memory and discarded. Nothing is logged or stored.

Open source

Frontend, backend, and training pipeline are all on GitHub. Audit everything.

Try it on a scan now.

Run the model against any lung CT slice. No signup, no tracking, no retention.

Open analyzer