README.md

metadata

license: cc-by-4.0
task_categories:
  - tabular-classification
  - tabular-regression
language:
  - en
tags:
  - cancer
  - oncology
  - synthetic
  - healthcare
  - sub-saharan-africa
  - pathology
  - histopathology
  - diagnosis
pretty_name: Cancer Pathology Histopathology Africa
size_categories:
  - 10K<n<100K
configs:
  - config_name: low_burden
    data_files: cancer_pathology_histopathology_low_burden.csv
  - config_name: moderate_burden
    data_files: cancer_pathology_histopathology_moderate_burden.csv
    default: true
  - config_name: high_burden
    data_files: cancer_pathology_histopathology_high_burden.csv
data_type: synthetic

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

Cancer Pathology & Histopathology Africa

Abstract

This synthetic dataset represents cancer pathology and histopathology data across sub-Saharan Africa, capturing diagnostic information, IHC marker availability, and turnaround times. The dataset contains 3,600-5,400 records per scenario with IHC marker availability varying by facility type (35-92%).

1. Introduction

1.1 Problem Statement

Pathology services are essential for cancer diagnosis and treatment planning, yet access varies dramatically across sub-Saharan Africa. National referral centers may have comprehensive IHC capabilities, while district facilities often lack basic immunohistochemistry. Turnaround times can be weeks in resource-limited settings.

1.2 Purpose

This dataset supports:

• Pathology service capacity assessment
• Diagnostic quality improvement initiatives
• Resource planning for pathology services
• Research on diagnostic delays

2. Methodology

2.1 Target Population

• Geographic scope: Sub-Saharan Africa
• Population represented: Cancer pathology specimens
• Time period: 2018-2025

2.2 Key Parameters

• IHC availability: National referral (92%), Regional (72%), District (35%)
• Turnaround time: 3-28 days
• Diagnosis confirmed: 85%
• Grade distribution: Well (12%), Moderate (35%), Poor (38%), Undifferentiated (15%)

2.3 Scenario Design

Scenario	Description	Records
low_burden	Higher resource setting	3,600
moderate_burden	Standard setting	4,500
high_burden	Lower resource setting	5,400

3. Dataset Description

3.1 Key Variables

• pathology_id, country, year, facility_type
• cancer_type, icd_o_morphology_code, morphology_description
• grade, lymphovascular_invasion, perineural_invasion
• margin_status, ihc_er, ihc_pr, ihc_her2, ihc_ki67_index
• molecular_subtype (breast), specimen_type
• diagnosis_confirmed, turnaround_time_days

3.2 Morphology Codes (ICD-O)

Code	Description	Frequency
8140/3	Adenocarcinoma	35%
8070/3	Squamous cell carcinoma	25%
8500/3	Ductal carcinoma	15%
8260/3	Papillary carcinoma	8%

4. Data Sources

• WHO Classification of Tumours (IARC)
• African pathology registry data
• College of American Pathologists guidelines
• Peer-reviewed literature on pathology in LMICs

5. Use Cases

• Pathology service capacity assessment
• IHC availability analysis
• Diagnostic turnaround time research
• Quality improvement initiatives

6. Citation

@dataset{cancer_pathology_histopathology_africa_2025,
  title={Cancer Pathology Histopathology Africa},
  author={Electric Sheep Africa},
  year={2025},
  publisher={HuggingFace},
  url={https://huggingface.co/datasets/electricsheepafrica/cancer-pathology-histopathology-africa}
}

7. License

Creative Commons Attribution 4.0 (CC-BY-4.0)