AI-natomy 101

AI-natomy 101

[Originally posted on Substack. Subscribe here to read it first.]

Lately, my inbox has been a trove of breakthroughs in AI for healthcare. We’re seeing AI-powered advances for Parkinson’s disease, Alzheimer’s disease, ophthalmological conditions, mental illnesseven triaging emergency room patients.

These leaps hint at a reality where healthcare is not only better but also more accessible. Eager to explore deeper, I connected with a few doctors-turned-founders of healthcare AI-powered nonprofits in our network and took their temperature on the state of medical AI. Their responses shed light on real-world applications and challenges. Ready to feast your A-eyes on the future?


The Remedy for the Data Divide

Ersilia was founded on the belief that 1) data science offers an unprecedented opportunity in biomedical research, and 2) all countries must benefit from it. Their tool, Ersilia Model Hub, is the largest collection of ready-to-use AI/ML models for infectious and neglected disease research, especially in the Global South. A recent Nature article co-authored by Ersilia's founders highlights a glaring data issue: most clinical trials have taken place in the Global North, creating a critical gap in our understanding of African genetics. Use of AI can indeed bridge this gap. However, the real focus should be on building the infrastructure, skills, and policies that empower local scientists to lead this transformation. When it comes to research, homegrown data is the best medicine.

+ The scarcity of unbiased, labeled data in low- and middle-income areas is one of the biggest topics in AI right now. There are many great resources and initiatives, including Tech Matters Better Deal for Data and Lacuna Fund .


"We know people are proximate to their problems, and for some reason, we’re okay with it. But we also know that we don’t do very much to help them be proximate to their solutions." - Vilas Dhar, President, Patrick J. McGovern Foundation        

Swapping the Sterile for the Sincere

At the heart of HERA Digital Health ’s current efforts is their Predictive Care Navigator, an AI-powered chatbot to converse with refugee women about their sexual and reproductive health needs. Named Amti, which means “aunty” in Arabic, HERA's chatbot is conversational, caring, and culturally appropriate — something that is missing in many healthcare chatbots. The research-backed tool integrates into WhatsApp, meeting users where they are and making it highly cost effective. Plus, the software engineer who coded and trained Amti is a refugee themself. HERA’s blend of medical accuracy, cultural empathy, and lived experience ensures every conversation feels more heart-to-heart than heart-to-hard drive.


"Medicine used to be the art of intuition, but now it's the science of information." - Eric Topol, Deep Medicine: How Artificial Intelligence Can Make Healthcare Human Again        

Healthcare Workers’ Right-hand Bot

Intelehealth is an open-source telemedicine platform connecting patients and frontline health providers in rural communities with remote doctors to deliver affordable healthcare. Powered by the AI digital assistant Ayu, the platform uses evidence-based protocols to improve patient outcomes. Ayu helps shift tasks from doctors to frontline health workers by recording patient history in the local language, translating it for the provider, and compiling it into a clinical summary. It also generates likely diagnoses, treatment pathways, and triage decisions, then translates this information back to the patient. Built on GPT-4 Turbo, Ayu has shown high diagnostic accuracy in controlled settings. Intelehealth is now refining Ayu to ensure its effectiveness in diverse environments. High-quality, affordable medical care for rural populations is just what the doctor ordered.


Quick Bytes

Other Sector Stories

APN Funding News


Let’s Talk

I am living and breathing AI for humanity these days. If you are too, let’s talk!


Link Library?

Miquel Duran-Frigola

Lead scientist and Co-Founder at Ersilia

6 个月

Great article, Kevin Barenblat! Thanks for highlighting Ersilia :)

要查看或添加评论,请登录

Fast Forward的更多文章