Featured paper: Towards Trustworthy Breast Tumor Segmentation in Ultrasound using Monte Carlo Dropout and Deep Ensembles for Epistemic Uncertainty Estimation
What if AI could admit when it's confused and help doctors catch cancer more safely? In this episode, we explore groundbreaking research on trustworthy breast tumor segmentation that flips the script on black-box AI. Discover how researchers uncovered shocking flaws in the popular BUSI dataset, duplicate images, jaw scans labeled as breasts, and why this "data leakage" made AI look far better than it actually was. Learn how Monte Carlo Dropout and Deep Ensembles teach AI to measure its own uncertainty, creating "heat maps" that highlight exactly where the model is struggling. We dive into why an AI that runs 25 times slower but admits confusion is actually safer for doctors, explore what happens when AI meets completely new, unfamiliar images, and unpack why this human-AI partnership could revolutionize breast cancer detection in low-resource settings. Join us as we investigate how teaching machines to say "I don't know" makes them more trustworthy, and ultimately more powerful tools for saving lives.

*Disclaimer: This content was generated by NotebookLM and has been reviewed for accuracy by Dr. Tram.*

Featured paper: Synthesizing scientific literature with retrieval-augmented language models

What if an AI could read 45 million scientific papers in seconds—and actually tell the truth about its sources? In this episode, we explore OpenScholar, a breakthrough retrieval-augmented language model designed to help researchers navigate the overwhelming flood of new scientific literature.

Learn why general-purpose AI models like GPT-4o hallucinate citations up to 90% of the time, how OpenScholar uses a unique iterative self-feedback loop to "fact-check" and refine its own answers, and why this fully open-source tool is outperforming multi-billion dollar proprietary systems. We dive into the OpenScholar DataStore (OSDS)—the largest open-access database of its kind—and discuss how this "super-librarian" AI is achieving expert-level accuracy that human PhDs actually prefer over 50% of the time.
*Disclaimer: This content was generated by NotebookLM and has been reviewed for accuracy by Dr. Tram.*

Featured paper: Quantification of Skeletal Muscle Perfusion in Peripheral Artery Disease Using 18F-Sodium Fluoride Positron Emission Tomography Imaging

What if a routine “bone scan” could reveal how well blood is actually reaching your leg muscles? In this episode, we explore a breakthrough study using 18F‑sodium fluoride PET to map muscle perfusion in peripheral artery disease. Learn why this scan outperforms the ankle‑brachial index for matching real symptoms, how it tracks recovery after revascularization, and why its “off‑the‑shelf” tracer could make advanced perfusion imaging more accessible.
*Disclaimer: This content was generated by NotebookLM and has been reviewed for accuracy by Dr. Tram.*