Datasets:

ClarusC64
/

clinical-organ-coupling-cascade-v0.1

scenario_id string	map_t0 int64	map_t1 int64	map_t2 int64	lactate_t0 float64	lactate_t1 float64	lactate_t2 float64	creatinine_t0 float64	creatinine_t1 float64	creatinine_t2 float64	spo2_t0 int64	spo2_t1 int64	spo2_t2 int64	platelet_t0 int64	platelet_t1 int64	platelet_t2 int64	fluid_response float64	ventilation_support int64	metabolic_noise float64	chart_noise float64	label int64
C001	74	73	72	2.1	2.2	2.3	1	1.1	1.2	96	95	94	210	205	200	0.72	0	0.41	0.32	0
C002	73	70	67	2.2	2.8	3.5	1.1	1.4	1.8	95	92	89	205	190	170	0.38	1	0.43	0.35	1
C003	75	74	74	1.9	2	2.1	0.9	1	1	97	96	95	215	210	205	0.78	0	0.39	0.31	0
C004	72	69	65	2.3	3	3.8	1.2	1.6	2.1	94	90	86	200	185	165	0.36	1	0.45	0.36	1
C005	74	73	73	2	2.1	2.2	1	1.1	1.2	96	95	95	208	204	200	0.74	0	0.4	0.33	0
C006	71	68	64	2.4	3.1	4	1.3	1.7	2.3	94	90	85	198	182	160	0.34	1	0.46	0.37	1
C007	76	75	74	1.8	1.9	2	0.9	1	1	97	97	96	218	214	210	0.8	0	0.38	0.3	0
C008	73	69	66	2.1	2.7	3.4	1.1	1.5	2	95	91	87	205	188	168	0.37	1	0.44	0.35	1
C009	75	74	73	1.9	2	2.1	0.9	1	1.1	97	96	95	215	210	205	0.79	0	0.39	0.31	0
C010	72	68	65	2.3	2.9	3.7	1.2	1.6	2.2	94	90	86	200	185	165	0.35	1	0.45	0.36	1
C011	74	73	72	2	2.1	2.2	1	1.1	1.2	96	95	94	210	205	200	0.76	0	0.41	0.32	0
C012	73	70	66	2.1	2.8	3.6	1.1	1.5	2	95	91	87	205	188	170	0.36	1	0.44	0.35	1
C013	76	75	75	1.8	1.9	2	0.9	1	1	97	97	96	220	215	212	0.82	0	0.37	0.3	0
C014	72	69	65	2.3	3	3.9	1.2	1.7	2.4	94	90	84	198	182	160	0.33	1	0.46	0.37	1
C015	75	74	73	1.9	2	2.1	0.9	1	1.1	97	96	95	215	210	205	0.79	0	0.39	0.31	0
C016	74	73	72	2.1	2.2	2.3	1	1.1	1.2	96	95	94	210	205	200	0.75	0	0.41	0.32	0
C017	72	69	65	2.2	2.9	3.7	1.2	1.6	2.2	95	90	86	202	186	165	0.35	1	0.45	0.36	1
C018	76	75	74	1.8	1.9	2	0.9	1	1	97	97	96	218	214	210	0.81	0	0.38	0.3	0
C019	73	69	66	2.1	2.8	3.5	1.1	1.5	2.1	95	91	87	206	190	168	0.36	1	0.44	0.35	1
C020	75	74	73	1.9	2	2.1	0.9	1	1.1	97	96	95	215	210	205	0.78	0	0.39	0.31	0
C021	71	67	63	2.5	3.2	4.2	1.4	1.8	2.6	93	88	83	195	178	155	0.32	1	0.47	0.38	1
C022	76	75	74	1.8	1.9	2	0.9	1	1	97	97	96	218	214	210	0.82	0	0.37	0.3	0
C023	73	69	66	2.2	2.8	3.5	1.1	1.5	2	95	91	87	205	188	170	0.37	1	0.44	0.35	1
C024	75	74	73	1.9	2	2.1	0.9	1	1.1	97	96	95	215	210	205	0.79	0	0.39	0.31	0
C025	72	68	64	2.4	3.1	4	1.3	1.7	2.3	94	89	85	198	180	158	0.34	1	0.46	0.37	1
C026	76	75	74	1.8	1.9	2	0.9	1	1	97	97	96	218	214	210	0.83	0	0.38	0.3	0
C027	71	67	63	2.5	3.3	4.3	1.4	1.9	2.7	93	88	82	195	176	150	0.31	1	0.47	0.38	1
C028	75	74	73	1.9	2	2.1	0.9	1	1.1	97	96	95	215	210	205	0.79	0	0.39	0.31	0
C029	72	68	65	2.3	3	3.8	1.2	1.6	2.2	95	90	86	200	185	165	0.35	1	0.45	0.36	1
C030	76	75	74	1.8	1.9	2	0.9	1	1	97	97	96	218	214	210	0.82	0	0.38	0.3	0

clinical-organ-coupling-cascade-v0.1

What this dataset does

This dataset evaluates whether models can detect instability arising from multi-organ coupling.

Each row represents a short trajectory across cardiovascular, metabolic, respiratory, renal, and hematologic indicators.

Core stability idea

Clinical collapse frequently occurs when stress signals across organs reinforce each other.

Examples include:

declining oxygenation combined with rising lactate
renal stress combined with metabolic deterioration
platelet decline combined with circulatory instability

These cascades are difficult to detect from single-variable thresholds.

Prediction target

label = 1 → instability due to multi-organ cascade
label = 0 → stable multi-organ trajectory

Row structure

Each scenario contains:

MAP trajectory
lactate trajectory
creatinine trajectory
oxygen saturation trajectory
platelet trajectory
fluid response indicator
ventilation support flag

Decoy variables:

metabolic_noise
chart_noise

These variables appear relevant but do not determine the label alone.

Evaluation

Predictions must use:

scenario_id,prediction C101,0 C102,1

Run evaluation:

python scorer.py --predictions predictions.csv --truth data/test.csv

Metrics returned:

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 ClarusC64 stability-reasoning benchmark family. Datasets share a latent stability geometry but expose only observable proxy variables.

Production Deployment

This dataset is intended as a compact benchmark for interaction-based instability detection.

Enterprise & Research Collaboration

The dataset supports research into cross-domain stability reasoning and multi-variable cascade detection.

License

MIT

Downloads last month: 16

Number of rows:

Total file size:

13.8 kB