Datasets:

ClarusC64
/

ffr-center-failure-surface-routing-v0.1

id stringclasses 6 values	site_id stringclasses 5 values	window_days int64 30 30	protocol_signature_hash stringclasses 5 values	motion_artifact_rate float64 0.05 0.18	signal_to_noise float64 0.7 0.88	plausibility_conflict_rate float64 0.04 0.2	rolling_auc float64 0.8 0.92	rolling_mae float64 0.05 0.11	rolling_mae_delta float64 0 0.06	calibration_error_shift float64 0 0.07	coherence_trend stringclasses 4 values	drift_type stringclasses 6 values	predicted_failure_risk float64 0.08 0.82	detection_confidence float64 0.78 0.88	drift_severity stringclasses 4 values	failure_surface_type stringclasses 6 values	subgroup_affected stringclasses 3 values	intervention_route stringclasses 6 values	recalibration_needed stringclasses 2 values	protocol_reset_plan stringclasses 5 values	escalation_priority stringclasses 5 values	confidence_score float64 0.81 0.9	notes stringclasses 6 values	constraints stringclasses 1 value	gold_checklist stringclasses 1 value
CFSR-001	SITE-UK-01	30	sigA1	0.05	0.88	0.04	0.92	0.05	0	0	stable	none	0.08	0.88	low	none	none	continue	no	none	none	0.9	baseline	<=320 words	surface+route+reset+priority+confidence
CFSR-002	SITE-EU-03	30	sigC2	0.14	0.75	0.1	0.87	0.08	0.03	0.03	down	protocol-drift	0.52	0.8	medium	protocol-drift-surface	broad	tighten-qa-and-monitor	yes	restore kernel + slice thickness	medium	0.84	early intervention	<=320 words	surface+route+reset+priority+confidence
CFSR-003	SITE-EU-03	30	sigC2	0.16	0.72	0.14	0.84	0.1	0.05	0.05	down-fast	silent-accuracy-decay	0.7	0.82	high	silent-decay-surface	broad	restrict-use-to-low-risk-cases	yes	freeze protocol; vendor audit	high	0.83	prevent audit failure	<=320 words	surface+route+reset+priority+confidence
CFSR-004	SITE-US-12	30	sigB9	0.16	0.72	0.18	0.83	0.09	0.04	0.06	down-fast	quality-driven-collapse	0.76	0.78	high	quality-collapse-surface	motion+low-snr	rescan-or-block-output	yes	motion control + HR management bundle	critical	0.81	block unsafe outputs	<=320 words	surface+route+reset+priority+confidence
CFSR-005	SITE-APAC-11	30	sigD5	0.18	0.7	0.2	0.8	0.11	0.06	0.07	down-fast	protocol-outside-envelope	0.82	0.8	critical	outside-envelope-surface	broad	disable-model-at-site	yes	protocol rollback to validated envelope	critical	0.86	stop deployment	<=320 words	surface+route+reset+priority+confidence
CFSR-006	SITE-US-07	30	sigB3	0.09	0.82	0.09	0.88	0.06	0.01	0.02	slight-down	minor-protocol-shift	0.22	0.82	low	minor-shift-surface	broad	tighten-qa-thresholds	no	none	low	0.86	small drift	<=320 words	surface+route+reset+priority+confidence

Goal

When a center drifts
route the safest operational response.

This dataset treats each site
as a failure surface.

Not a single bad scan.

Inputs

Rolling site window metrics:

image quality proxies
protocol signature
rolling performance (AUC, MAE)
calibration error shift
plausibility conflict rate
drift type and risk

Required outputs

failure_surface_type
subgroup_affected
intervention_route
recalibration_needed
protocol_reset_plan
escalation_priority
confidence_score

Failure surfaces

Examples:

minor shift surface
protocol drift surface
silent decay surface
quality collapse surface
outside envelope surface

Intervention routes

Examples:

continue
tighten QA and monitor
restrict use to low risk cases
rescan or block output
disable model at site

Why it matters

The same model can be safe at one site
and unsafe at another.

Hospitals need a clear route.
Vendors need early containment.

This dataset makes those actions
testable and repeatable.

Evaluation

The scorer checks the response includes:

surface label
subgroup label
route + reset plan
escalation priority
confidence score 0 to 1

Downloads last month: 22

Number of rows:

Total file size:

8.61 kB