Clinical decision-support prototype for validation. Physician confirmation required. Do not enter identifiable patient information.
CPE-001 · Acceptance fixtures

Synthetic Validation Library

15 fully de-identified synthetic cases. Use them to demonstrate clinical patterns or validate engine behavior against the canonical CPE-001 rule set.

Run validation dashboard
FX-001
Complex
Overt Primary Pattern

Overt primary hypothyroidism with autoimmune support

Recognize overt biochemical pattern and separate autoimmune etiology from functional state.

Biochemical Pattern Compatible with Overt Primary HypothyroidismPR-HT-01
Load case
FX-002
Complex
Euthyroid Restraint Pattern

Antibody positive with euthyroid function

Prevent treatment/functional failure classification from antibody positivity alone.

No Biochemical Evidence of Current Thyroid Dysfunction from Entered TSH and FT4PR-HT-06
Load case
FX-003
Complex
Subclinical Pattern

Mild subclinical pattern with positive Anti-TPO

Recognize possible subclinical pattern while prohibiting automatic therapy.

Biochemical Pattern Compatible with Possible Subclinical HypothyroidismPR-HT-02
Load case
FX-004
Standard
Subclinical Pattern

TSH at or above 10 with normal FT4

Avoid labeling severe TSH elevation as overt when FT4 remains within range.

Biochemical Pattern Compatible with Possible Subclinical HypothyroidismPR-HT-02
Load case
FX-005
High
Context-Sensitive Pattern

Pregnancy context with mild TSH elevation and Anti-TPO positivity

Route to pregnancy-specific pathway and prevent universal TSH >2.5 rule.

Pregnancy-Specific Thyroid Risk Context IdentifiedPR-HT-061 alert
Load case
FX-006
High
Subclinical Pattern

Preconception infertility context

Route guideline-sensitive fertility case without automatic treatment.

Biochemical Pattern Compatible with Possible Subclinical HypothyroidismPR-HT-021 alert
Load case
FX-007
Complex
Central Thyroid Axis Dysfunction

Central hypothyroidism after traumatic brain injury

Detect safety-critical central pattern despite normal TSH.

Possible Central Hypothyroidism Pattern — Safety Review RequiredPR-HT-031 alert
Load case
FX-008
Complex
Central Thyroid Axis Dysfunction

Low FT4 with inappropriately mild TSH elevation and pituitary history

Prevent missed central dysfunction when TSH is not low.

Possible Central Hypothyroidism Pattern — Safety Review RequiredPR-HT-031 alert
Load case
FX-009
Complex
Thyrotoxicosis Differentiation

Thyrotoxicosis with Anti-TPO positivity

Prevent premature hashitoxicosis diagnosis.

Thyrotoxicosis Pattern — Etiology Requires DifferentiationPR-HT-051 alert
Load case
FX-010
High
Overt Primary Pattern

Postpartum overt hypothyroid pattern with iron deficiency context

Detect overt pattern while routing postpartum context.

Biochemical Pattern Compatible with Overt Primary HypothyroidismPR-HT-011 alert
Load case
FX-011
Standard
Screening Trigger Only

LDL and CK pattern without thyroid labs

Use metabolic/muscular findings as screening trigger only.

Thyroid Evaluation Screening Trigger IdentifiedPR-HT-04
Load case
FX-012
High
Thyrotoxicosis Differentiation

Biotin use with discordant apparent thyrotoxicosis labs

Surface assay interference review.

Apparent Thyrotoxicosis Pattern — Assay Interference Review RequiredPR-HT-052 alerts
Load case
FX-013
Complex
Safety-Critical Endocrine Pattern

Severe decompensation risk alert

Escalate severe clinical pattern without routine outpatient framing.

Biochemical Pattern Compatible with Overt Primary HypothyroidismPR-HT-01PR-HT-071 alert
Load case
FX-014
High
Overt Primary Pattern

Amiodarone-associated overt biochemical pattern

Detect functional pattern and show medication etiology modifier.

Biochemical Pattern Compatible with Overt Primary HypothyroidismPR-HT-011 alert
Load case
FX-015
Standard
Euthyroid Restraint Pattern

No biochemical thyroid dysfunction with depression-like symptoms

Demonstrate restraint when symptoms lack thyroid biochemical pattern.

No Biochemical Evidence of Current Thyroid Dysfunction from Entered TSH and FT4
Load case
All cases are synthetic and de-identified. They are used only for software validation and are not derived from real patient encounters.