Mission
Last Week in Health AI curates the latest and most impactful AI in medicine research from top-tier journals and preprint servers. Our goal is to help researchers, clinicians, and healthcare professionals stay current with AI innovations transforming healthcare.
Sources
We aggregate research from multiple authoritative sources:
- PubMed/MEDLINE: Top medical journals including Nature, Science, JAMA, NEJM, The Lancet, BMJ, and specialized AI in medicine journals
- arXiv: Computer science and quantitative biology preprints
- bioRxiv & medRxiv: Biological and medical preprints
Curation Process
1. Multi-Source Aggregation
We query PubMed, arXiv, bioRxiv, and medRxiv weekly for papers involving AI/ML methods applied to medical problems.
2. AI-Powered Relevance Filtering
Each paper is evaluated by Gemini 2.0 for:
- Presence of AI/ML methods (deep learning, computer vision, NLP, etc.)
- Clear medical or healthcare application
- Technical rigor and novelty
- Clinical impact potential
Only papers scoring ≥0.65 on relevance are included.
3. Expert-Level Summarization
Curated papers receive comprehensive AI-generated summaries including:
- Executive summary for quick understanding
- Detailed key points covering methodology and results
- AI methods and medical context
- Clinical significance and impact assessment
- Limitations and future directions
Quality Assurance
- Strict relevance thresholds ensure high signal-to-noise ratio
- Multi-source aggregation provides comprehensive coverage
- Direct links to original papers for verification
- AI summaries are supplementary - original papers remain authoritative
Technology
- AI Engine: Google Gemini 2.0 Flash
- Data Sources: PubMed E-utilities, arXiv API, bioRxiv/medRxiv APIs
- Infrastructure: Python, GitHub Actions, GitHub Pages
- License: MIT (fully open source)
About the Curator
Bryan Tegomoh created this platform to bridge the gap between cutting-edge AI research and practical medical applications.
Contact: bryan@lastweekinhealthai.blog
GitHub: Repository
Disclaimer
Summaries are AI-generated for educational purposes only. Content may contain errors. Always consult original papers and qualified healthcare professionals for medical decisions. This site is not affiliated with PubMed, arXiv, or any publisher.