SignalSift

Turning a disposable blood sensor into a high-confidence Alzheimer's screening tool

Turning a disposable blood sensor into a high-confidence Alzheimer’s screening tool

SignalSift sits inside Agilus Neuro’s blood-test device and makes a low-cost disposable sensor perform like a far more expensive lab instrument. It automatically spots and discards unreliable readings, so the final result is one the clinician can trust.

Request information

The problem it solves

Faint signals can be drowned out by ordinary interference.

The signals that flag Alzheimer’s in a drop of blood are extremely faint. On a cheap disposable sensor, ordinary interference — electrical noise and proteins sticking to the sensor surface — can easily drown out that faint signal. The result: some healthy people get flagged as positive, a false positive.

NeuroFilter-AI fixes this in software. The device uses several tiny sensors at once; SignalSift checks every one, throws out the ones that are misbehaving, and calculates the answer only from the sensors it trusts. Fewer false positives, more confidence, no more-expensive hardware.

Conceptual signal correction

Illustrative NeuroFilter-AI signal filtering diagram

Illustrative figure: faint blood biomarker signals can be obscured by interference; NeuroFilter-AI helps remove unreliable sensor readings before the screening result is calculated.

How it works

Three steps before a screening answer is trusted.

Measure many times over

The device reads the same sample on several small sensors at once, giving multiple independent opinions instead of one.

Filter out bad readings

SignalSift runs quality checks on each sensor and uses AI to spot outliers — the sensors distorted by interference — and removes them.

Deliver a trustworthy result

It calculates the final answer from only the reliable sensors, and can tell the device to re-measure if quality is too low.

Key benefits

Smarter software for more reliable blood-based screening.

Fewer false positives

Rejects untrustworthy data so healthy people are less likely to be mis-flagged.

Lab-grade confidence, low-cost hardware

Gets high-end performance out of an inexpensive, disposable sensor.

Built-in quality control

Every result is automatically vetted; weak results trigger a re-test instead of a wrong answer.

Smarter, not just bigger

An AI layer that improves accuracy without changing the physical test.

Designed for point-of-care

Works from a tiny blood sample, suited to clinics rather than only central labs.

Within the Agilus Neuro platform

Software quality control for Alzheimer’s biomarker screening workflows.

SignalSift minimizes bias and variance in concentration measurements by multiple biosensors in the same blood plasma sample, thereby directly improving the classification of Alzheimer’s disease biomarkers.

Patent pending

SignalSift is a product of Agilus Neuro, LLC. Patent pending. SignalSift is intended to support biomarker screening and is not a stand-alone diagnostic. Any figures shown are illustrative. Product information is subject to change.

Agilus Neuro

Biosensor platforms for biomarker detection, spun out of Washington University in St. Louis.

info@agilusneuro.com