FIGO AUB System 1
The international classification for abnormal uterine bleeding by frequency and regularity, with thresholds banded by age.
Clinical reference: Munro et al., 2018.
Clinical engine
Published criteria, applied precisely.
Shiora's classification engines are deterministic: a documented rule, a citable threshold and a result a clinician can reproduce.
Standards implemented
The criterion remains visible.
Deterministic classification makes every output auditable, publishable and defensible to a regulator.
Published standard
STRAW+10
The reference criteria for staging the menopausal transition, computed from period-start dates.
Clinical reference: Harlow et al., 2012.
PHQ-9 and GAD-7
Widely used depression and anxiety measures, scored to published thresholds and tracked against cycle phase.
EPDS
The Edinburgh Postnatal Depression Scale, an international standard for perinatal mood screening.
Sexual and pelvic function measures
Validated measures scored to their published algorithms, in the same record as everything else.
Why determinism matters
A rule can be inspected, reproduced and cited.
A model must be validated, monitored for drift and explained. A deterministic rule can show exactly which date and threshold produced a classification.
That separation is deliberate. It keeps language generation away from the point of clinical classification.
Clinical integrity
The engine knows when a criterion does not apply.
Where medication, pregnancy, surgery or a condition makes a classification invalid, Shiora explains why instead of returning a result that looks confident and is wrong. Knowing when not to answer is a core engineering decision.
See it in context