Bio-RAMP Labs

This study is either going to be a breakthrough moment or a spectacular failure for applying #LLMs to healthcare in Africa.🤞 Today on the #podcast, we sit down with the team running the largest study ever done on how LLMs like #ChatGPT will - and won't - work for healthcare in Africa. Tobi Olatunji MD, Founder and CEO of Intron Health, speaks with Bilal A Mateen, Executive Director of Digital Square at PATH to discuss the upcoming AfriMed-QA project which will evaluate 20 LLMs over 32 clinical specialties across 15 countries in Africa. This work was made possible through the support of Google Research, Bill & Melinda Gates Foundation, and Digital Square at PATH. Implementing partners include Masakhane, SisonkeBiotik, and Bio-RAMP Labs. 🎧 Tune in wherever you get your podcasts https://lnkd.in/dxiXSPnk 🩺 If you're an African clinician, follow the link in the comments if you'd like to contribute to the study. ✍️ Drop your name in the comments if you'd like to receive a summary of the results once they're available. Be patient. It's going to take a while. #NLP #AIforGood Leah Ekbladh #digitalhealth

Teaming up with Google, Bill & Melinda Gates Foundation, Digital Square at PATH and inviting all African clinicians to contribute to the largest ever study on LLMs (like ChatGPT, Gemini) in African healthcare [across 15 countries and 32 specialties]. Working with incredible partners like SisonkeBiotik, Bio-RAMP Labs, and Masakhane, the project will create a large (20k ) Africa-focused, indigenously sourced, uniquely human, Medical Question-Answering (QA) benchmark dataset to rigorously evaluate multiple LLMs in the African context, elicit their strengths and weaknesses, identify high-value use cases, quantify their risk of bias and harm, and finetune more culturally attuned LLMs for deployment in African clinics. Just like doctors won't prescribe new drugs without rigorous testing, shiny tech should be put through similar rigor before exposing patients and providers to unknown risks. As these tools get better and more pervasive, care needs to be taken so they are deployed where they work well, and limited where they fail. Right now, across LMICs, no one knows! Intron Health wants to help close this gap. Over 1,000 clinicians and non-clinicians across 15 countries have already started contributing Africa-specific questions, answers, MCQ/SAQ Exam questions, simulated doctor-patient conversations, summaries, translations, and model ratings. Through grants from Google Research, Bill & Melinda Gates Foundation, and PATH, we’ll pay a modest stipend to appreciate your time and effort!! Please share to all African clinicians in your network to help us spread the word. Visit https://lnkd.in/eme_GY28 to indicate interest and start contributing today. #LLMs #innovation #digitalhealth #ai #aihealth #naturallanguageprocessing #nlp #medicine #clinicalresearch #technology #healthtech #aiforgood Chris Fourie Mercy Nyamewaa Asiedu, Ph.D. Charles Nimo Melissa Miles Bilal A Mateen Sewu-Steve T., Laura Adams Rowena Luk Maria Rotilu Jeff Dean Andrew Bredenkamp, PhD. Emmanuel Adegboye Ato Bentsi-Enchill Dayo Koleowo Dr. Dotun Olowoporoku Desigan Chinniah Bayo (Olubayo) Adekanmbi PhD Lola Walker Graham Walker, MD Spencer Dorn Matthew Lungren MD MPH Roxana Daneshjou Ajibade Ashaye, MD MBA Sion Hau Julianne Roseman Ekin Günaysu Kirsten Connell Timnit Gebru Edgar Camelo Mohamed Nanabhay