Researchers have introduced two new benchmarks, mamabench and mamaretrieval, designed to evaluate retrieval-augmented generation (RAG) systems specifically for maternal, neonatal, and reproductive health. These benchmarks address a gap in existing medical QA datasets by focusing on the unique queries of nurse-midwives and providing a chunk-level relevance benchmark for maternal health guidelines. Additionally, a companion paper details MAM-AI, an on-device RAG system for nurse-midwives in Zanzibar that operates fully offline, demonstrating that even small, on-device models can achieve competitive retrieval performance. AI
IMPACT These benchmarks and the MAM-AI system could improve the accessibility and accuracy of medical information for healthcare professionals in resource-limited settings.
RANK_REASON The cluster consists of two research papers introducing new benchmarks and an associated on-device AI system for a specific medical domain.
Read on arXiv cs.IR (Information Retrieval) →
- alphaXiv
- arXiv
- CatalyzeX
- DagsHub
- Gotit.pub
- HealthBench
- Hugging Face
- mamabench
- mamaretrieval
- maternal, neonatal, and reproductive health
- ScienceCast
- Gemma
AI-generated summary · Google Gemini · from 4 sources. How we write summaries →
