Why Your Kidneys Might Be Tricking Your Brain (And How Glassbury AI Fixes It)
If a patient's kidneys aren't clearing waste efficiently, p-tau217 levels in the blood naturally back up and spike—creating a terrifying "false positive" for Alzheimer’s pathology in patients who are cognitively healthy.
At Glassbury AI, we believe true precision medicine requires algorithms that understand basic biology—including the plumbing.
The Problem: When Biology Mimics Pathology
For clinicians, the rise of p-tau217 as a blood-based biomarker for Alzheimer’s disease has been revolutionary. But biomarkers don't exist in a vacuum. If a patient has Chronic Kidney Disease (CKD), their kidneys may struggle to clear waste efficiently. This leads to an accumulation of plasma proteins, including p-tau217, in the bloodstream.
The result? A "false positive" spike that suggests neurodegeneration, when in reality, the patient's brain health is perfectly fine.
The Solution: Normalizing the Noise
We’ve developed a multi-input regression model at Glassbury AI designed to filter out this renal noise.
Our algorithm doesn't just look at raw p-tau217 concentrations. It integrates critical patient data, including:
eGFR (estimated Glomerular Filtration Rate)
Cystatin C levels
By learning the mathematical relationship between renal clearance decay and plasma protein accumulation, our machine learning model dynamic-shifts the diagnostic threshold based on an individual's kidney health.
