Acute Coronary Instability Predictor (ACIP)

Interactive 24–48 hour clinical risk simulator using a shared global pattern model plus personalised N-of-1 adaptation.

Demonstration only. This simulator uses synthetic coefficients and illustrative thresholds. It is not a validated medical device and must not be used for diagnosis or patient care.

Patient and current state

ACIP uses the same adaptive learning platform and infrastructure which enables existing predictors such as Falls, Sepsis and COPD, centred on Medibiosense wearable monitors.
ACIP uses the platform's existing medication module.

Background risk

Illustrative 1–9 scale

Physiology: current versus personal baseline

Personal baseline: 68 bpm
Personal baseline: 38 ms
Personal baseline: 92 mmHg
Personal baseline: 15/min
Personal baseline: 97%
100% = usual personal activity

Personal baseline settings

Rhythm and medication/context flags

Predicted short-horizon instability

0–24 hour risk
—
24–48 hour risk
—
Current tier
—
Change from last cycle
—
Illustrative trajectory signal

Why the score changed

Largest model drivers

Global + N-of-1 mathematics

Personal deviation:
zi,x(t) = [xi(t) − μi,x] / σi,x

Global model:
logit(pglobal) = β0 + Σ βkfk(t)

Personalised output:
logit(pi) = αi + γ · logit(pglobal)

The global component represents patterns learned across a cohort. The N-of-1 layer interprets current physiology relative to this patient's own baseline and recalibrates the shared estimate.

Simulated six-cycle risk trajectory

Clinical interpretation

Demonstrator response:
Synthetic demonstrator Coefficients, thresholds and scenario trajectories are illustrative only. A clinical ACIP would require retrospective development, prospective validation, calibration testing, external validation, clinical safety assessment and regulatory review.
Clinical concept, model and design: Michael Tremblay.
Intellectual property owner: Cassis Ltd. Copyright © 2026 Cassis Ltd. All rights reserved.
This demonstrator, including its software, interface, text, visualisations and associated materials, is protected by copyright and other applicable intellectual property rights. No licence for reproduction, adaptation, redistribution or commercial use is granted except with prior written permission from Cassis Ltd.