Mercy Nyamewaa Asiedu, Ph.D
@dr_nyamewaa
Followers
650
Following
630
Media
29
Statuses
285
Research Scientist, Google Research. Co-founder, @gaphealthtech. Passionate about using tech to democratize health.
San Francisco, CA
Joined September 2016
Excited to share a project I’m extremely proud of, our publicly available health benchmark dataset, AfriMed-QA, which spans 60 medical schools, 30 specialties and 16 countries across Africa. The paper won the best social impact paper award at ACL 2025 in Vienna.
Introducing AfriMed-QA – the first large-scale pan-African dataset designed to help evaluate & develop optimized and effective LLMs for African healthcare.
4
29
67
Many thanks to amazing collaborators @tobiolatunji_ @AbrahamOwos @moorekwesi @amilah_dul @aka_chineme @Chris_Fourie_SA @bonadossou @yuehgoh and others 😊
0
0
6
This work was in collaboration with @IntronHealth @SBiotik @MasakhaneNLP @gatesfoundation @path , the Federation of African Medical Students Association, University of Cape Coast, @GeorgiaInstitu and many more
0
0
0
Excited to share a project I’m extremely proud of, our publicly available health benchmark dataset, AfriMed-QA, which spans 60 medical schools, 30 specialties and 16 countries across Africa. The paper won the best social impact paper award at ACL 2025 in Vienna.
Ensuring generalization of LLMs in response to distribution shifts is especially important for medical and health-related models. Here we describe AfriMed-QA, an open-source benchmark question–answer dataset sourced from countries across Africa. More at https://t.co/Z7vsK5jKRC
0
1
3
Congratulations to the authors of "AfriMed-QA: A Pan-African, Multi-Specialty, Medical Question-Answering Benchmark Dataset", recipient of the Best Social Impact Award at #ACL2025!
1
5
24
A Pan-African medical Question-Answer (QA) dataset designed to evaluate and develop equitable and effective LLMs for African healthcare. @googleafrica @GoogleAI @dr_nyamewaa @tobiolatunji_
https://t.co/MZ7CvatDFN
research.google
0
2
8
Many thanks to all colleagues who contributed to this work: @kat_heller, @weballergy, @ChintanGhate, @adoubleva, Oluwatosin Akande, @gsiwo, @SteveAdudans, @SylvanusAitkins, @odia_Ehi_ , and @Makuto4
0
1
6
How do large language models(LLMs) perform on out-of-distribution health datasets? Do tropical and infectious diseases serve as an out-of-distribution case to test LLMs? Does performance demonstrate potential use for global health surveillance? Check out our blog to learn more🦟
Today we announce TRINDs, a dataset and benchmarking pipeline that uses synthetic personas to train and optimize performance of LLMs for tropical and infectious diseases, which are out-of-distribution for most models. Learn more → https://t.co/R1PFhQ0c8R
1
8
13
Google Research is proud to contribute to Google's inaugural Health Report, reaffirming our commitment to human-centered AI for global health. Explore the full report to learn more:
📢 New: Google's Impact on Health Explore how Google is advancing AI, delivering trusted health information, transforming health organizations, and building a thriving global health ecosystem. 📖 Dive into the report: https://t.co/wFjmhL05vE
48
72
247
Many thanks to all co-authors: @kat_heller, @weballergy, @ChintanGhate, @adoubleva, Oluwatosin Akande, @gsiwo, @SteveAdudans, @SylvanusAitkins, @odia_Ehi_ , and @Makuto4
0
4
8
If you are at @NeurIPSConf this week you might have caught our TRINDs demo at the @GoogleAI Booth. If you didn't, come check out our posters at the GenAI4health( https://t.co/rlJL9uInfe) and AIM-FM( https://t.co/hjjIfgc5YF) workshops :)
0
0
3
We find that including context such as risk factors and location in addition to symptoms also improves model performance. Additionally we assemble a panel of human experts to set a human expert baseline score on the dataset and to provide ratings of data quality, usefulness, etc
1
0
1
Results show that LLMs perform worse on TRINDs than they do on reported US-based health QAs, indicating a distribution shift in the data and the need for further optimization for global diseases. We also find that LLMs more accurately identify diseases that are common or specific
2
1
2
We develop and expand the TRopical and INfectious Diseases (TRINDs) dataset to evaluate LLMs for these contexts, and demonstrate through systematic experimentation, the effect of contextual information on LLM outputs for disease classification.
1
0
1
Check out our latest work on Contextual Evaluation of Large Language Models for Tropical and Infectious Diseases ( https://t.co/CVzyk9EgDS), accepted at two NeurIPS workshops: GenAI4health ( https://t.co/rlJL9uInfe) and AIM-FM ( https://t.co/hjjIfgc5YF).
1
9
25