The Dr. Hyman Show

We Can Detect Cancer Years Earlier—So Why Aren’t We?

4 snips
Apr 22, 2026
Guest
Daniel K. Sodickson
Dr. Daniel K. Sodickson, physicist and imaging pioneer who helped advance MRI tech, explains why shifting from one‑time scans to tracking over time could catch disease earlier. He discusses using baseline MRIs, longitudinal imaging to cut false positives, AI plus prior data for faster cheaper scans, and how imaging complements blood tests for proactive health monitoring.
Ask episode
AI Snips
Chapters
Books
Transcript
Episode notes
INSIGHT

Imaging As Proactive Longitudinal Surveillance

  • Medical imaging can shift from a one-time diagnostic snapshot to proactive longitudinal surveillance that detects disease earlier.
  • Daniel K. Sodickson argues baselines and repeat scans reveal change over time, reducing late-stage diagnoses like advanced cancer.
INSIGHT

Context Cuts Imaging False Positives

  • False positives from imaging are use-dependent, not device-fixed, and drop when images are interpreted in context.
  • Feeding prior scans and clinical data into models cut false positives from ~64% to below 10% in prostate cancer prediction.
ADVICE

Get A Baseline MRI And Build Context

  • Establish a personal baseline MRI and pair it with blood tests and genetics to create actionable context.
  • Use that baseline as a reference for future scans so changes trigger informed follow-up instead of alarm.
Get the Snipd Podcast app to discover more snips from this episode
Get the app