---
license: cc-by-4.0
task_categories:
- image-classification
- text-classification
- other
language:
- en
- fr
- ar
- sw
- ha
- yo
- am
- wo
- tw
- zu
- af
tags:
- medical
- xray
- fracture
- bone
- africa
- multimodal
- healthcare
- radiology
- synthetic-data
size_categories:
- 1K<n<10K
dataset_info:
  features:
  - name: record_id
    dtype: string
  - name: patient_id
    dtype: string
  - name: image
    dtype: image
  - name: fracture_type
    dtype: string
  - name: age
    dtype: int32
  - name: gender
    dtype: string
  - name: country
    dtype: string
  - name: region_type
    dtype: string
  - name: primary_language
    dtype: string
  - name: chief_complaint
    dtype: string
  - name: complaint_language
    dtype: string
  - name: heart_rate
    dtype: int32
  - name: blood_pressure_systolic
    dtype: int32
  - name: blood_pressure_diastolic
    dtype: int32
  - name: temperature
    dtype: float32
  - name: height
    dtype: float32
  - name: weight
    dtype: float32
  - name: pain_scale
    dtype: int32
  - name: season
    dtype: string
  - name: phone_type
    dtype: string
  - name: digital_literacy
    dtype: string
  splits:
  - name: train
    num_bytes: 45000000
    num_examples: 790
  - name: validation
    num_bytes: 6500000
    num_examples: 113
  - name: test
    num_bytes: 13000000
    num_examples: 226
  download_size: 64500000
  dataset_size: 64500000
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train.jsonl
  - split: validation
    path: data/validation.jsonl
  - split: test
    path: data/test.jsonl
---

# African Medical Multimodal Bone Fracture Dataset

## Dataset Description

This dataset is a comprehensive, multimodal bone break classification dataset specifically designed for African healthcare contexts. It addresses critical gaps in medical AI for resource-constrained environments while ensuring cultural sensitivity and local relevance.

### Dataset Summary

- **Total Records**: 1,129 multimodal medical cases
- **Original Images**: 1,128 X-ray images (89% of dataset)
- **Augmented Images**: 135 mildly augmented images (11% of dataset)
- **Fracture Types**: 10 different bone fracture classifications
- **Countries Represented**: 18 African countries
- **Languages**: 11 African languages + English
- **Modalities**: 6 data types per record

### Key Features

🌍 **African Context Integration**
- Geographic diversity across 18 African countries
- Rural (59%), Urban (26%), Peri-urban (15%) distribution
- Cultural factors including traditional medicine usage
- Socioeconomic indicators relevant to healthcare access

🗣️ **Multilingual Clinical Data**
- Patient histories in 11 languages: English, French, Arabic, Swahili, Hausa, Yoruba, Amharic, Wolof, Twi, Zulu, Afrikaans
- Cultural pain expressions and symptom descriptions
- Traditional medicine integration patterns

🏥 **Realistic Equipment Simulation**
- 89% high-quality images (representing modern African hospitals)
- 11% mildly augmented images (rural/portable equipment variations)
- Quality scores: 6.2-9.5/10 (realistic for African healthcare)

📱 **Mobile Health Integration**
- Smartphone usage patterns and digital literacy levels
- GPS tracking and movement data
- SMS/voice follow-up preferences

## Usage Examples

### Loading the Dataset

```python
from datasets import load_dataset

# Load the dataset
dataset = load_dataset("electricsheepafrica/african-medical-multimodal-fracture")

# Access splits
train_data = dataset["train"]
val_data = dataset["validation"]
test_data = dataset["test"]

# View a sample
sample = train_data[0]
print(f"Patient from {sample['country']} with {sample['fracture_type']}")
print(f"Complaint: {sample['chief_complaint']} ({sample['complaint_language']})")
print(f"Vitals: HR {sample['heart_rate']}, BP {sample['blood_pressure_systolic']}/{sample['blood_pressure_diastolic']}")
```

### Accessing Images

```python
from PIL import Image
import requests

# Get image path and load
image_path = sample['image_path']
# Images are stored as relative paths in the dataset
```

## Fracture Types

The dataset includes 10 fracture classifications:

| Fracture Type | Description |
|---------------|-------------|
| Fracture Dislocation | Joint displacement with bone break |
| Comminuted fracture | Multiple bone fragments |
| Pathological fracture | Break due to disease/weakness |
| Avulsion fracture | Bone fragment pulled by tendon/ligament |
| Greenstick fracture | Incomplete break (common in children) |
| Hairline Fracture | Thin crack in bone |
| Spiral Fracture | Twisting break pattern |
| Oblique fracture | Diagonal break across bone |
| Impacted fracture | Bone ends driven together |
| Longitudinal fracture | Break along bone length |

## Geographic Distribution

### Countries Represented
Nigeria, Kenya, Ghana, South Africa, Tanzania, Uganda, Ethiopia, Morocco, Egypt, Senegal, Rwanda, Botswana, Zambia, Zimbabwe, Mali, Burkina Faso, Cameroon, Ivory Coast

### Regional Distribution
- **Rural**: 665 cases (59%) - Reflects African population distribution
- **Urban**: 293 cases (26%) - Major cities and towns
- **Peri-urban**: 171 cases (15%) - Transitional areas

## Clinical Measurements

Each record includes realistic clinical measurements:
- **Heart Rate**: 60-100 bpm (normal range)
- **Blood Pressure**: Systolic 90-140 mmHg, Diastolic 60-90 mmHg
- **Temperature**: 36.0-38.5°C
- **Height**: 150-190 cm
- **Weight**: 45-90 kg
- **Pain Scale**: 1-10 (patient-reported pain level)

## Ethical Considerations

### Synthetic Data Approach
- All patient data is **synthetically generated** to protect privacy
- No real patient information is included
- Cultural patterns are based on research and literature, not individual cases

### Cultural Sensitivity
- Traditional medicine practices are represented respectfully
- Cultural pain expressions reflect documented patterns
- Family involvement in medical decisions is appropriately modeled

## Applications

### Primary Use Cases
- **Multimodal Medical AI**: Training models that integrate images, text, and structured data
- **Cross-cultural Healthcare**: Developing culturally-aware diagnostic systems
- **Resource-constrained Settings**: AI systems for rural and underserved areas
- **Medical Education**: Training materials for African healthcare contexts

## Citation

If you use this dataset in your research, please cite:

```bibtex
@dataset{african_medical_multimodal_2024,
  title={African Medical Multimodal Bone Fracture Dataset},
  author={ElectricSheepAfrica},
  year={2024},
  publisher={Hugging Face},
  url={https://huggingface.co/datasets/electricsheepafrica/african-medical-multimodal-fracture}
}
```

## License

This dataset is released under the Creative Commons Attribution 4.0 International License (CC BY 4.0). You are free to use, modify, and distribute this dataset for any purpose, including commercial use, as long as you provide appropriate attribution.

---

*This dataset represents a comprehensive effort to create culturally-aware, multimodal medical AI training data for African healthcare contexts. While synthetic, it is based on extensive research and aims to promote more inclusive and effective medical AI systems.*