Upload a CT scan
The system converts the scan to a 3-D volume, extracts the lungs (also from full-body CT), finds nodules, turns each one into a row of measurements and runs XGBoost to estimate whether it is a lung-cancer nodule. You get a report a doctor can act on.
🔒 Please upload anonymised scans only (no patient names or IDs). Your scan and results are deleted automatically after 2 hours. Research prototype, not a medical device.
or try a sample:
- 13-D volumeresampled to 1 mm voxels
- 2Lung extractionauto-crops lungs from full-body scans
- 3Nodule detectionmulti-scale 3-D morphology
- 4Feature table539 LungLocAI-3D features + 33 shape measures
- 5XGBoostLungLocAI-3D model (AUC 0.97) + cancer probability
- 6Doctor reportPDF + Lung-RADS-style triage
Results
Impression
Computer-generated decision support from a research prototype. A radiologist must confirm every finding on the original images.
CT slices
3-D lungs & nodules
drag to rotateHigh Intermediate Low Not graded