HemodynamiCARDIOMETAIQ / IN-SILICO TRIAL EXPLORATION
Explore a virtual cohort.
Inspect the assumptions.
Define a synthetic population. Compare treatment and control. See how eligibility, response variability and the number of simulated profiles affect the modelled readout.
RESEARCH DEMONSTRATION · SYNTHETIC DATAWhat this demonstration does: analytic weight-loss and responder simulations using synthetic profiles. It is a separate research model from the homepage trajectories. Clinical efficacy, patient-specific accuracy and trial-enrolment requirements are not established by this demonstration.
SIMULATED READOUT
Semaglutide vs control
Week 52 · 200 profiles per arm · Seed 7
Results match the design shown.
Weight-loss distribution
Virtual participants in each weight-loss range. Negative values indicate weight gain.
SemaglutideControl
View distribution values
| Weight-loss range (%) | Treatment count | Control count |
|---|---|---|
| -20.0 to -15.8 | 0 | 0 |
| -15.8 to -11.7 | 0 | 2 |
| -11.7 to -7.5 | 3 | 5 |
| -7.5 to -3.3 | 13 | 32 |
| -3.3 to 0.8 | 16 | 44 |
| 0.8 to 5.0 | 37 | 51 |
| 5.0 to 9.2 | 48 | 29 |
| 9.2 to 13.3 | 34 | 27 |
| 13.3 to 17.5 | 23 | 8 |
| 17.5 to 21.7 | 17 | 2 |
| 21.7 to 25.8 | 6 | 0 |
| 25.8 to 30.0 | 3 | 0 |
The eligible virtual cohort
525profiles screened
61.9%screen failure
67.7mean age (years)
36.6mean BMI (kg/m²)
90.0%with modelled T2DM
Weight-loss responders
| Threshold | Treatment | Control |
|---|---|---|
| ≥ 5% | 65.5% | 33.0% |
| ≥ 10% | 38.5% | 14.5% |
| ≥ 15% | 17.5% | 2.0% |
Exploratory subgroups
| Group | N / arm | Model difference (pp) |
|---|---|---|
| Without T2DM | 20 | 11.1 |
| With T2DM | 180 | 5.0 |
| BMI ≥ 40 | 41 | 5.1 |
| BMI < 40 | 159 | 5.7 |
These are descriptive summaries of synthetic outputs. They do not quantify real-world treatment efficacy, clinical prediction error or the required size of a clinical trial. Subgroup differences are exploratory and reflect the model’s assumptions.
What this workspace runs
This page adapts the virtual-population and analytic weight-loss trial workflow in the supplied CardioMetaIQ application. It uses seeded sampling, correlated baseline features and a synthetic lifestyle-like control. It is separate from the v4.8 ODE trajectories on the homepage.
Generate & screen
Correlated virtual baseline profiles are screened by age, BMI, E/e′ ≥ 13 and NYHA II–III. The same eligible profiles underpin both arms; this is a simulated comparison, not recruitment or randomisation of real participants.
Compare outcomes
The source analytic model estimates weight-loss responses with variability, non-response and residual noise. Positive values mean weight loss; negative values mean weight gain. Diabetes status attenuates the model’s treatment response.
Understand the limits
Increasing virtual sample size reduces sampling variation within this model; it does not establish clinical validity. This page reports descriptive outcomes only. A real study requires independently justified assumptions and a prespecified statistical analysis plan.
No patient data are collected. Results are not evidence of drug efficacy, personalised treatment recommendations or a substitute for a prespecified statistical analysis plan. The source app’s full ODE, imaging, operational forecasting and trial-intelligence modules are not executed on this page.