MedVision Edge โ Offline Chest X-ray Analysis
AI-powered chest X-ray screening using Gemma 4 E4B fine-tuned on NIH ChestX-ray14 (~23K training samples with 5x oversampling).
Detects 5 pathologies: Pneumonia, Consolidation, Cardiomegaly, Pleural Effusion, Pulmonary Edema
Validated on two independent benchmarks
| Pathology | Base AUC | Fine-tuned AUC | CheXpert (Gold Std) | ฮ vs Base |
|---|---|---|---|---|
| Cardiomegaly | 0.490 | 0.832 | 0.723 | +70% |
| Pleural Effusion | 0.605 | 0.703 | 0.797 | +16% |
| Pulmonary Edema | 0.688 | 0.753 | 0.668 | +9% |
| Consolidation | 0.599 | 0.627 | 0.667 | +5% |
| Pneumonia | 0.519 | 0.617 | 0.501* | +19% |
Base AUC: unmodified Gemma 4 (zero-shot). Fine-tuned AUC: our model, evaluated on 1,103 held-out NIH images. CheXpert: same model evaluated on 500 independent images with 5-radiologist consensus labels (Stanford).
*Pneumonia: insufficient CheXpert prevalence (2.2%). Detection under active development.
AI screening tool only. Not for clinical diagnosis. All findings must be confirmed by a qualified radiologist.
Output Language
Example chest X-rays โ Expected results
CheXpert test set, radiologist-verified
1. NormalAll clear โ no pathology detected
2. CardiomegalyCardiomegaly: DETECTED
3. EffusionEffusion: DETECTED
4. EdemaEdema: DETECTED
5. MultiplePneumonia: DETECTED
Consolidation: DETECTED
Effusion: DETECTED
Consolidation: DETECTED
Effusion: DETECTED
Click an example to load it
| Upload Chest X-ray | Output Language | Patient Age (years) | Patient Weight (kg) |
|---|
About MedVision Edge
- Model: Gemma 4 E4B-it, fine-tuned with Unsloth QLoRA (r=64, 82M trainable params)
- Training data: NIH ChestX-ray14 (112,120 image dataset), ~23K training samples with 5x oversampling and augmentation
- Evaluation: NIH test set (1,103 images) + CheXpert gold standard (500 images, 5 radiologist consensus)
- Protocols: WHO IMCI 2024 clinical guidelines (deterministic function calling, zero hallucination)
- Languages: 140+ supported natively by Gemma 4
- Deployment: Runs offline on consumer GPU (5GB GGUF via Ollama) or this Gradio demo
- GPU used: ~43h on NVIDIA RTX 5070 Ti 16GB (training + evaluation)
Born at the Gemma 4 Good Hackathon | Apache 2.0