featured_project.ipynb
Stroke Prevalence Analysis
Screening data from the Amuwo Odofin community, Lagos
An independent, end-to-end analysis of a real community health-screening dataset — built as a reproducible package, not a one-off script.
Method
- Cleaned and structured raw screening data
- Applied WHO 2000 BMI classification
- Applied ACC/AHA 2017 hypertension thresholds
- Applied WHO 2008 waist-to-hip ratio standards
- 19-section exploratory & statistical analysis
Delivered as
- Jupyter Notebook (full analysis)
- Written portfolio report
- README + requirements.txt + .gitignore
- Fixed a real Python 2.7 / 3.14 environment conflict (MGLTools) via PATH ordering
Key finding — Using a composite stroke-risk score, 76.7% of the 60 adults screened fell into the moderate-to-high risk band (40% high, 36.7% moderate) — with 71.7% also showing high cardiovascular risk by waist-to-hip ratio. In a single community screening, more than three in four adults carried elevated stroke risk.
Figure — from the notebook
