Oncologist builds CT reading demo with a 4B open model that boxes lung nodules
Scobleizer · x · 2026-10-31
An oncologist demoed a small tool built on a 4-billion-parameter open model for chest CT reading: ask a question in plain language and the model answers while highlighting the exact region it looked at — typing "find the nodules in the lungs" successfully located and boxed a nodule.
It can also compare two scans to answer whether lesions are growing, shrinking, or stable — the key question in oncology follow-up. The author stresses it's a research demo, not a medical device: sensitivity/specificity untested, and a radiologist still reviews every scan. Inference takes about 10 seconds.
More from Multimodal
- Hotel room scanned in 4 minutes, Echo-3 reconstructs it perfectly — risingodegua · 2026-10-31
- AI short film made in 12 hours with open-source Longshot Studio, full project files released — R34vspec · 2026-10-31
- AI film festival Lumara premieres its 74-minute lineup free on X — Scobleizer · 2026-10-31
- PhobosStream: Run 14B Video DiTs on 8GB GPUs With 788MB Peak VRAM via Disk Streaming — martianacademy · 2026-10-31
- 8 AI video models, one prompt: Seedance 2.5 vs Kling 4.0 head-to-head showdown — WorldofAI · 2026-10-31
- Halloween horror video generated on just 8GB VRAM with LTX 2.5 + LoRA — izzray · 2026-10-31