An AI system that identifies physiological disorder in ICU patients up to 10 hours before cardiac arrest — validated on 1,310 confirmed cases from MIMIC-IV.
Existing early warning systems like MEWS and NEWS2 are rule-based threshold tools. They wait for a value to breach a limit. By then — it's often too late.
The real signal is not the breach. It's the disorder building toward it.
Current systems detect deterioration after significant physiological decline has already occurred
Threshold-based alarms generate high false positive rates causing alarm fatigue in clinical staff
No existing system measures biological signal order — the structural coherence that collapses before catastrophic events
Hundreds of thousands of in-hospital cardiac arrests annually — most preceded by detectable physiological signals
Unlike systems that monitor absolute values, Regenesis-OS measures the structural organization of biological signals over time. As a system approaches collapse — disorder emerges.
Continuous vital sign data ingested from existing hospital monitoring infrastructure. No new hardware required.
Raw signals transformed into spatiotemporal signal fields capturing relationships across time, space, and biological subsystems.
Custom metrics quantify deviation from baseline signal order — entropy, variance, gradient energy, anisotropy, cross-system coupling.
Temporal sequences analyzed to detect and classify five distinct disruption modes with associated confidence values.
Closed-loop system selects interventions based on detected pattern and confidence — modulating intensity to biological response.
Regenesis-OS classifies the type of biological disorder — not just its presence
USPTO — Systems and Methods for Detecting and Restoring Biological Signal Order Using Adaptive Closed-Loop Control
The same disorder signal that precedes cardiac arrest precedes every major physiological collapse event. Regenesis-OS is a platform.
Deploy in ICU environments as a software layer on existing monitoring infrastructure. Clinical validation. FDA pathway.
Expand to sepsis, respiratory failure, renal failure, hepatic collapse. One platform predicting all major in-hospital deterioration events.
Wearable integration detects pre-Alzheimer, pre-cancerous, and metabolic disorder years before clinical diagnosis.
Continuous whole-body biological stability monitoring. The intelligence layer inside every hospital bed and home device on earth.
Justin C. Radford-Berry spent seven years as an independent quantitative researcher — finding anomalous patterns in complex, noisy, real-time data streams.
He noticed something: the methodology for detecting market disorder and the methodology for detecting biological disorder are structurally identical. Both are entropy detection problems in time series data.
He turned that lens on ICU telemetry data — and found a signal nobody had measured before.
Biological systems don't fail suddenly. They warn you. Rising entropy. Increasing variance. Loss of signal organization. The disorder precedes the collapse — every single time. We built the system that listens.
Detect biological collapse before it happens — giving clinicians the time they need to intervene, and giving patients the chance they deserve.
Whether you're an investor, clinician, researcher, or potential partner — we want to hear from you.