Medical AI Update: From Computer Predictions to Real Patient Results
By Mark C. Engasser, MD — Board-Certified Orthopedic Surgeon
Artificial intelligence is beginning to move beyond identifying promising drugs and interpreting medical images. We are now seeing encouraging evidence of how AI-assisted discoveries may benefit actual patients.
Two recent developments deserve attention.
1. AI-Assisted Drug Discovery Shows Promise in Preventing Colorectal Cancer
On November 2, Recursion Pharmaceuticals reported encouraging Phase II results for REC-4881, an experimental medication being studied for familial adenomatous polyposis (FAP).
FAP is an inherited condition that causes numerous intestinal polyps and carries a very high lifetime risk of colorectal cancer.
Among 15 patients evaluated after 12 weeks of treatment:
- 87% experienced a reduction in polyp burden.
- The median reduction in polyp burden was 47%.
- Some patients maintained their improvement after treatment ended.
AI-enabled biological screening helped identify this treatment opportunity.
Why is this exciting?
Patients with FAP often undergo repeated colonoscopies and may eventually require major surgery. A medication that safely reduces polyp formation could potentially change how this condition is managed.
However, this was a small, uncontrolled study. We do not yet know whether the medication reduces cancer risk, prevents surgery, or provides lasting benefits.
2. AI Finds Heart Valve Abnormalities That May Otherwise Go Unreported
Another study, presented November 1, evaluated AI software designed to identify aortic valve calcification on routine chest CT scans.
The system demonstrated approximately 98% sensitivity and 97% specificity in a study involving 440 patients.
Interestingly, many scans showing calcification had no explicit mention of the finding in the original radiology report.
Why does this matter?
Aortic valve calcification can be associated with aortic stenosis, a condition that may eventually require valve replacement.
AI could help physicians recognize potentially important findings on imaging studies already performed for other reasons.
The technology remains investigational, however, and detecting calcification does not necessarily mean a patient has clinically significant valve disease.
Dr. Engasser's Perspective
As an orthopedic surgeon, I find one aspect of medical AI particularly fascinating: We are moving from AI that simply provides information toward AI that may actually improve medical decisions and treatment.
The most important question is no longer whether AI can identify abnormalities or suggest promising medications.
It is whether these discoveries will ultimately reduce illness, prevent unnecessary procedures, improve quality of life, and help people live healthier lives.
That is where the real medical revolution will occur.
And it brings us back to a theme I frequently discuss:
The Longer You Live, the Longer You Will Live.
Continued medical innovation may allow more people to remain healthy long enough to benefit from the next generation of discoveries. It is an exciting possibility—not a guarantee.
The bottom line: These two developments are encouraging, but neither has yet demonstrated improved long-term patient outcomes. The next step is larger clinical trials that establish safety and meaningful benefits.
Sources and Further Reading
- Recursion Pharmaceuticals, November 2, 2026Company website and clinical research updates
- Nanox Imaging, November 2026Company website and AI imaging research
Mark C. Engasser, MD
Board-Certified Orthopedic Surgeon
Doctor Pickleball
Medicine. Innovation. Healthy Living. Generosity.
Dr. Pickleball's Musings reflects personal observations and opinions intended to encourage thoughtful discussion.