Smart Medicine

About Smart Medicine

Smart Medicine explores how AI, physics, and human insight are reshaping the future of healthcare. Hosted by Dr. Ali Fatemi, a medical physicist and entrepreneur, the show dives into the latest innovations in medical imaging, oncology, and digital health. Each episode brings conversations with clinicians, scientists, and innovators to uncover how new technologies are transforming patient care — and what challenges still remain. Whether you’re a healthcare professional, researcher, startup founder, or simply curious about the future of medicine.

Publisher

Dr. Ali Fatemi

Category

health & fitness

Language

en

Explicit

No

First episode

6 Oct 2025

Latest episode

5 Oct 2026

Latest episodes

33 episodes in the feed.

1. 

   5 Oct 2026

   Why 90% of Radiology AI Fails in Clinical Practice (and How to Fix It) ft. Dr. Woojin Kim (Radiologist)

   Is AI actually making radiologists faster, or is it just creating more pop-up windows?In this episode of Smart Medicine, Dr. Ali Fatemi sits down with Dr. Woojin Kim—MSK radiologist, former CMIO at Rad AI, CMO of the ACR Data Science Institute, and co-founder of three successful healthtech companies.They dive deep into why billions of dollars have been poured into clinical AI, yet real-world adoption remains surprisingly low.Key Discussion Highlights: 📍 The 2 Biggest Gaps in Radiology AI: Why fancy standalone viewers fail and why benchmark accuracy (like 95% PPV) collapses when applied to low-prevalence real-world clinical data. 📍 Clinical Utility: Model Accuracy: Health system executives don't just care about model numbers; they care about workflow integration, throughput, ROI, and clinical utility. 📍 Building a #FrontierPractice: How health systems must approach AI by starting with real pain points, dirty data governance, and culture change rather than just buying software. 📍 Will AI Replace Radiologists? Why this is the wrong question when imaging volume grows 4–6% annually while workforce growth remains stagnant. 📍 Project Presto in Action: How Dr. Kim's team at Hopper uses vision-language models (VLMs) to auto-populate existing dictation templates💬 Question for you: What is the single biggest workflow friction point in your clinical practice right now? Let us know in the comments below! 👇🔔 Subscribe to Smart Medicine for more deep dives into AI, medical physics, and clinical practice!

2. 

   22 Sept 2026

   AI in Medical Physics: Innovation, Job Evolution & Global Collaboration | Smart Medicine Podcast

   Welcome to the Smart Medicine Podcast hosted by Dr. Ali Fatemi! In this episode, senior medical physicist Dr. Jihong Wang from MD Anderson shares his perspective on AI in medical imaging, the need for clinical validation, healthcare leadership, and why international scientific collaboration is vital to future medical breakthroughs.⏱️ Timestamps & Key Topics:00:00 - AI in Medical Imaging & Spectroscopy: Multi-dimensional data analysis (T1, T2, DWI, DCE) and automating prostate spectroscopy workflows.03:15 - Why Validation is Critical: Deep learning reconstruction risks, soft tissue artifacts, and FDA alerts.08:00 - Will AI Replace Medical Physicists & Physicians?: Task automation (contouring, coding) vs. expanding clinical capacity.11:30 - Healthcare Leadership & Vision: Leading by example and supporting early-career team members.15:00 - Global Scientific Collaboration: Why healthcare must transcend national boundaries, referencing lessons from early 20th-century physics.19:00 - Mentorship & Career Advice: Finding key mentors and learning how to learn.23:00 - The Future of Medicine: Upcoming "Penicillin moments" in cancer biology and the promise of FLASH radiotherapy.🔔 Subscribe & Connect: If you enjoyed this episode, please like, subscribe, and click the Notification Bell for more conversations on medical technology, radiation oncology, and healthcare leadership!💬 Question for the Community: How are AI tools currently impacting your clinical or research workflow? Let us know in the comments below!#MedicalPhysics #RadiationOncology #SmartMedicinePodcast #AIinMedicine #FLASHtherapy #Radiology #MedicalImaging #Mentorship

3. 

   8 Sept 2026

   Why Medical AI is Solving the Wrong Problem (and How to Fix It) | Smart Medicine Ep. 1 🩺💻

   AI is the hottest topic in medicine, but why are so many clinical AI tools failing to deliver a real ROI?In this premier episode of the Smart Medicine Podcast, host Dr. Ali Fatemi (clinical medical physicist) sits down with Dr. Khan Siddiqui—a rare "quadruple threat": a practicing radiologist, academic researcher, software engineer, and deep-tech entrepreneur.Dr. Siddiqui has spent his career at the absolute cutting edge of technology. He learned to code in the 6th grade on a PDP-11 gamma camera system in Pakistan, built epidemiological linear regression models for the CDC in 1993, and went on to work at Microsoft, where his team's machine learning work laid the foundation for the Xbox Kinect, Azure ML, and the revolutionary ResNet neural network architecture. Later, he served as Chief Medical Officer and Chief Strategy Officer at Hyperfine, co-inventing the world's first portable bedside MRI. Today, he is the founder and CEO of Hopper, building multimodal foundation models to streamline clinical workflows.In this episode, Dr. Siddiqui breaks down why improving lesion detection is NOT the key to clinical AI success, explains why AI will never replace radiologists, and shares how we can build AI that actually saves clinicians from crushing burnout.⏱️ TIMESTAMPS TO WATCH:00:00 – Introduction & Dr. Siddiqui's remarkable background.02:15 – Coding in the 6th grade & building linear regression models for the CDC in 1993.06:30 – Choosing radiology and pioneering the world's first imaging informatics fellowship.12:00 – The Microsoft years: Random forests, Xbox Kinect, Azure ML, and the birth of ResNet.16:15 – Redefining medical imaging: The story behind Hyperfine's portable bedside MRI.21:30 – The Windows Kernel Story: "Treating a complex engineering bug like a difficult patient".24:00 – Why clinical AI tools fail: The 250-millisecond detection myth vs. workflow reality.33:00 – What is driving the crushing rise in global imaging volume?.40:30 – Demystifying Multimodal AI & vector representations in medical imaging.48:00 – Hopper & Presto: Creating an "ambient scribe" that integrates directly into existing templates.54:30 – Will AI replace radiologists? Jevons Paradox, horse carriages, and the autopilot analogy.59:00 – Dr. Siddiqui's personal "Why": Helping his overworked radiologist wife reclaim her evenings.1:02:15 – Personalized medicine, AlphaFold, and losing 130 lbs using data and LLMs.💡 KEY TAKEAWAY FROM THE EPISODE: "You always innovate at the edge of your experience." To solve healthcare's biggest challenges, we must stop building isolated AI "point solutions" and start building generalized models that adapt directly to clinicians' existing workflows.Subscribe for more deep dives into the future of smart medicine and medical technology! 🔔

Contact Smart Medicine

Guest appearances

Books guests

Based on episode analysis; this does not confirm that the show is currently accepting guests.