Levels

@Levels

Levels helps you see how food affects your health.

Levels
@Levels
Sep 8
More health data could mean fewer unnecessary tests. A nodule seen once can trigger a CT or biopsy. Ten years of data showing it never changed tells a very different story. Ziad Obermeyer explains why longitudinal measurement can help separate signal from noise.
00:00
1
Levels
@Levels
Sep 7
Ziad Obermeyer says healthcare is entering a “mainframe to PC” transition. More biological data can now be collected cheaply outside hospitals. Add AI that can interpret it continuously, and the center of healthcare may start moving toward the patient. #AI #wearables
00:00
1
Levels
@Levels
Sep 6
Why doesn’t heart disease create the same urgency as cancer? Dr. Ami Bhatt says patients are often told they have risk factors, then left to manage them quietly for years. By the time medicine truly mobilizes, disease may already be established.
00:00
3
Levels
@Levels
Sep 4
Most medical AI is trained to copy doctors. Ziad Obermeyer’s team trained an algorithm to learn from patients instead, and it found knee pain radiologists were missing. His point: AI gets much more interesting when existing doctors stops being the answer key. #AI #medicalAI
00:00
Levels
@Levels
Sep 3
Every year, 400,000 people drop dead of a heart attack - many weren’t known to be at risk. Ziad Obermeyer’s AI found a high-risk group using ECG data, and 85% weren’t flagged by current clinical criteria. The warning signal was already there. Medicine hadn’t learned to see it.
00:00
4