File size: 1,089 Bytes
4fff1c8
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
---
language:
- en
---
# 🦴 OsteoLogic: Clinical Fracture Detection

OsteoLogic is a custom computer vision model designed for radiographic fracture detection. It uses a Keras-based architecture and is deployed with `.safetensors` for secure, memory-efficient inference.

## 🔍 Model Details
- **Input shape**: 224×224×3
- **Classes**: Fracture / No Fracture
- **Format**: `.safetensors`
- **Framework**: Keras

## 🚀 Usage
Upload a radiograph via the web interface. The model will preprocess, predict, and return a visual overlay with classification.

## 🧠 Deployment Philosophy
This model is part of a doctrine-driven pipeline focused on:
- Reproducibility
- Minimal UI friction
- Full memory restoration
- Version clarity

## 📦 Files
- `osteologic.safetensors`: Model weights
- `config.json`: Architecture metadata
- `app.py`: Gradio interface
- `requirements.txt`: Pinned dependencies
- `runtime.txt`: Python version for Hugging Face Spaces

## 🧪 Clinical Context
Designed for orthopedic screening workflows. Not intended for diagnostic use without clinical oversight.