AI Is Reading 15 Million X-Rays a Year With No Human in the Loop | Prashant Warier, Qure.ai

What you'll learn

• You learn how Qure.ai autonomously reads 15 million X-rays per year with no human radiologist in the loop, across 70 countries
• You see the CREATE study results, where the algorithm correctly identifies 54 of 100 high-risk lung cancer patients versus 2 of 100 in standard CT screening
• You understand why AI can detect lung cancer more than 60 days earlier than conventional diagnostic pathways
• You discover how the system analyzes routine X-rays people already get for unrelated reasons, without adding new screening burden
• You hear the prediction that within 5 to 10 years primary care will be AI-first, with the first conversation happening with AI instead of a doctor

Frequently asked questions

Topics

Read next

• Screenpoint Medical helpt radiologen borstkanker eerder opsporen met AI

  Het Nijmeegse Screenpoint Medical ontwikkelt AI-software die radiologen ondersteunt bij het lezen van mammogrammen. Het product Transpara is inmiddels in meer dan 30 landen in gebruik en heeft ruim 12 miljoen mammogrammen verwerkt.

• Nederlandse startups die AI inzetten voor medische beeldvorming zoeken hun weg door certificering en ziekenhuismarkten

  Een cluster Nederlandse startups ontwikkelt AI-software voor röntgen, MRI, pathologie en mammografie. De weg naar CE-markering en integratie in ziekenhuisworkflows blijkt daarbij minstens zo bepalend als de technologie zelf.

• Dertig uur zoeken naar één incident

  Dertig uur lang camerabeelden doorscrollen voor één incident: een frustrerende realiteit in de zorg. Dit verhaal vormde de basis voor Oddity.ai’s innovatieve AI-oplossing.

• Kaiko bouwt een AI-assistent die meewerkt in de dagelijkse oncologiepraktijk

  Het Amsterdamse Kaiko ontwikkelt kaiko.w, een AI-assistent voor oncologieteams die werkt met klinische tekst, medische beeldvorming en genomische data. Sinds 2021 gebruikt het Nederlands Kankerinstituut het systeem in de dagelijkse praktijk.

• De arts die AI bouwt, en de Nederlandse zoektocht naar durf

  De Nederlandse zorg kraakt onder administratieve druk, met bijna de helft van de medewerkers die een kwart van hun tijd kwijt is aan papierwerk. Een arts pakt deze uitdaging aan door zelf AI te bouwen en zoekt naar de durf om dit systeem te transformeren.

Description from the channel

80% of lung cancer cases are diagnosed too late, not because the signals aren't there, but because the system wasn't designed to catch them at the right moment. Qure.ai is changing that. In this episode, Craig Smith sits down with Prashant Warier, co-founder and CEO of Qure.ai, to explore how the company has built the most scaled autonomous AI use case in healthcare today, reading 15 million chest X-rays per year across 70 countries with no radiologist in the loop. The conversation centers on the CREATE study, which validated Qure.ai's Lung Nodule Malignancy Risk Score: a system that analyzes routine X-rays people already get for unrelated reasons, identifies high-risk nodules, and flags which patients need follow-up CT scans. The results are striking: standard CT screening programs find cancer in about 2 out of 100 patients; Qure.ai's algorithm identifies 54 out of 100 high-risk patients who turn out to be positive on CT. Warier also shares his prediction that within 5 to 10 years, primary care will be AI-first, meaning the first conversation you have when something feels wrong will be with an AI, not a doctor. Key Topics Covered: ● The CREATE study: how Qure.ai's Lung Nodule Malignancy Risk Score improves positive cancer detection from 2 out of 100 to 54 out of 100 on follow-up CT ● Why 80% of lung cancer is diagnosed late, and how the signals are already present in routine imaging that millions of people get every year ● The most scaled autonomous AI use case in healthcare: 15 million TB X-rays per year, no human radiologist in the loop, across 70 countries ● Why AI can detect lung cancer more than 60 days earlier than conventional diagnostic pathways ● The regulatory and practical barriers to making AI diagnostics directly available to patients, and why it still has to go through the medical establishment ● Warier's 5-to-10-year prediction: primary care will be AI-first, with patients talking to AI before they ever see a doctor Lung cancer is the world's deadliest cancer almost entirely because it's found too late, and Qure.ai's approach of using imaging that already exists, rather than adding new screening burden, is one of the most scalable paths to changing that outcome globally. Subscribe to Eye on A.I. for weekly conversations with the people building and deploying the future of AI. Craig Smith on X: https://x.com/craigss EYE On A.I. on X: https://x.com/EyeOn_AI Connect with Prashant Warier LinkedIn: https://in.linkedin.com/in/pwarier Timestamps 00:00 How AI Is Transforming Healthcare Diagnostics 02:17 The Origin Story of Cure.ai 04:03 Moving From Image Recognition to Early Disease Detection 05:02 Why AI in Healthcare Requires FDA Approval 06:31 How Cure.ai Combines Vision Models and LLMs 07:56 Inside the Radiologist Workflow With AI 10:38 The Global Impact of AI-Powered Tuberculosis Screening 12:17 How AI Improves Accuracy and Early Detection 14:08 The CREATE Study and Lung Cancer Screening Breakthroughs 19:31 How Cure.ai Is Expanding Across Healthcare Systems 21:34 Beyond Lung Cancer: Heart Disease, Stroke & Fractures 25:07 The Future of AI-Powered Diagnostics 28:11 Why Earlier Detection Will Define the Future of Healthcare 31:39 Should Patients Have Direct Access to AI Diagnostics? 36:19 Will AI Become the Primary Care Doctor of the Future? 38:22 The Most Scaled Autonomous AI System in Healthcare Today