ClarusC64/clinical-microvascular-instability-v0.1

clinical-microvascular-instability-v0.1

What this dataset does

This dataset evaluates whether models can detect instability in microvascular circulation.

Each row represents a simplified microcirculatory monitoring scenario observed across three time points.

The task is to determine whether microvascular flow remains stable or is moving toward microvascular instability.

Core stability idea

Microcirculatory stability depends on interaction between capillary flow distribution and tissue oxygen extraction.

Signals that interact include:

• —capillary flow proxy trajectory
• —oxygen extraction proxy trajectory
• —microvascular density proxy
• —lactate trajectory
• —tissue metabolic demand
• —intervention delay

Instability emerges when capillary flow heterogeneity rises while tissue extraction and lactate increase.

Prediction target

label = 1 → microvascular instability label = 0 → stable microcirculation

Row structure

Each row includes:

• —capillary flow proxy trajectory
• —oxygen extraction proxy trajectory
• —microvascular density proxy trajectory
• —lactate trajectory
• —tissue demand proxy
• —intervention delay

Decoy variables:

• —lab_noise
• —chart_noise

Evaluation

Predictions must follow:

scenario_id,prediction

Example:

MV101,0 MV102,1

Run:

python scorer.py --predictions predictions.csv --truth data/test.csv --output metrics.json

Metrics produced:

accuracy precision recall f1 confusion matrix dataset integrity diagnostics

Structural Note

This dataset reflects latent stability geometry through observable proxies.

The generator and latent rule structure are not included.

This dataset is part of the Clarus Stability Reasoning Benchmark.

License

MIT