---
license: cc-by-4.0
task_categories:
- tabular-classification
language:
- en
tags:
- digital-health
- artificial-intelligence
- clinical-decision-support
- CDSS
- synthetic
- sub-saharan-africa
pretty_name: AI & Clinical Decision Support
size_categories:
- 10K<n<100K
configs:
- config_name: ai_deployed
  data_files: data/ai_cdss_ai_deployed.csv
- config_name: ai_pilot
  data_files: data/ai_cdss_ai_pilot.csv
  default: true
- config_name: ai_absent
  data_files: data/ai_cdss_ai_absent.csv
data_type: synthetic
---

> ⚠️ **Synthetic dataset** — Parameterized from published SSA literature, not real observations. Not suitable for empirical analysis or policy inference.

# AI & Clinical Decision Support

## Abstract

Synthetic dataset modeling AI and clinical decision support system deployment across three SSA scenarios. Captures AI domains (TB CXR, retinopathy, triage), deployment modes, recommendation adherence, diagnostic accuracy, provider trust, validation status, barriers, and safety metrics. Parameterized from SSA AI health research.

## Parameterization Evidence

| Parameter | Value | Source | Year |
| --- | --- | --- | --- |
| Most SSA AI projects | Pilot stage | Wahl et al. BMJ Glob Health | 2018 |
| CDSS improves accuracy | 10-30% | Topol. Nat Med | 2019 |
| AI TB detection sensitivity | 90-98% | Mollura et al. Radiology | 2020 |
| Local validation critical | Bias concerns | Mollura et al. | 2020 |

## Validation

![Validation Report](validation_report.png)

## Usage

```python
from datasets import load_dataset
ds = load_dataset("electricsheepafrica/ai-clinical-decision-support", name="ai_pilot")
df = ds['train'].to_pandas()
```

## References

1. Wahl B et al. (2018). AI global health. *BMJ Glob Health*. DOI: 10.1136/bmjgh-2018-000798
2. Topol EJ (2019). High-performance medicine. *Nat Med*. DOI: 10.1038/s41591-018-0300-7
3. Mollura DJ et al. (2020). AI radiology LMICs. *Radiology*. DOI: 10.1148/radiol.2020201434

## License

CC-BY-4.0