This AI Urine Test Detects 50 Diseases and Reads What Genetic Tests Miss | #193This AI Urine Test Detects 50 Diseases and Reads What Genetic Tests Miss | #193
The Dr. Joy Kong Podcast
Dr. Joy Kong chats with Dayan Nenov about an AI-driven urine test that reads metabolic patterns to flag many diseases from a single sample. The conversation covers current clinical work, regulatory routes, and how AI might reshape diagnostics and everyday medical practice.
38:11•30 Jul 2026
Can One Urine Test Flag 50 Diseases? AI, Metabolomics and the Future of Diagnosis
Episode Overview
- Luventix uses AI and gas chromatography on urine to classify disease patterns across tens of thousands of data points without preselecting biomarkers.
- Clinical work on colorectal cancer, Crohn’s disease, coeliac disease and SIBO suggests urine-based metabolomic patterns can indicate diverse conditions.
- Metabolomics is presented as a real-time picture of what the body is doing, complementing genetics, which shows predisposition rather than current state.
- AI tools can save clinician time by summarising medical records, suggesting possible diagnoses and highlighting treatments that may otherwise be missed.
- The long-term goal is a low-cost urine panel that screens for many conditions at once, making broad early detection feasible for large populations.
“The most precious resource we have is time. It’s doctors and nurses’ hours in the day that they can see patients and actually do medicine and not fill out forms.”
What drives someone to seek a life without late diagnoses and guesswork in medicine? This conversation between Dr. Joy Kong and technology entrepreneur Dayan Nenov zeroes in on a bold idea: using artificial intelligence and a single urine sample to flag dozens of diseases before you feel unwell. Dayan shares how a TV segment on dogs smelling urine and sitting down when they detected cancer sparked the concept for his company, Luventix.
He and his co-founder asked a simple question: "How can we replicate this result with technology?" From there, they built a system that treats a dog’s nose and brain like high‑tech tools—a gas chromatograph and a classifier—and replaced them with machine learning that reads 38,250 data points from one urine sample. You’ll hear about their IRB‑approved clinical work on colorectal cancer, Crohn’s disease, coeliac disease and SIBO, and why they don’t chase individual biomarkers.
Instead, they let AI learn the patterns across nearly 10,000 metabolites in urine, giving a holistic snapshot of how an organism is behaving right now. Dr. Kong even notes how this feels like a bridge between Eastern systems thinking and Western lab science. The chat widens out to big questions about AI in healthcare: reducing costs, freeing doctors from endless paperwork, and helping clinicians spot conditions and treatments they might otherwise miss.
Dayan is clear that genetics show what we’re "programmed" for, while metabolomics shows what the body is actually doing in real time, and he sees real power in combining the two. If you’re curious how AI might one day turn a simple pee test into a low‑cost panel for dozens of conditions, and why early detection could become routine rather than rare, this episode offers a grounded look at where things stand today and where they might go next.
How would your choices change if a cheap, quick test could flag trouble months or years before symptoms hit?
Do you want to link to this podcast?
Get the buttons here!
More From This Show
The latest episodes from the same podcast.