07 / Embodied + recorded history
A snapshot is a window,
not the patient.
Five deidentified MIMIC-IV Demo trajectories show exactly which records were available at a shifted ICU cutoff. A four-lab frame stays separate from the longer sequence that precedes it.
01 / Exact bounded result
What was actually built
The Explorer verifies one exact source release before rendering. Every count below comes from the committed artifact, not from a live API.
02 / Cohort navigator
Five separate identifier scopes
Choose an alias to inspect its focus admission, latest ICU stay, bounded frame, and available prior records. Native subject_id, hadm_id, and stay_id never cross between aliases.
03 / Snapshot versus trajectory
One cutoff, two honest views
Loading history
04 / Source-returning timeline
Every shown event keeps its row locator
The newest matching events appear first. Prescription entries remain order-interval records; procedure entries remain recorded codes.
05 / Similar frame, different history
P04 and P05 are nearest only under one declared metric
A normalized four-lab distance of Loading makes this the closest pair in the five-frame cohort. That does not establish the same patient, clinical equivalence, or any future conclusion.
06 / Pinned evidence
The compact projection stays traceable to eight upstream files
Only the declared record families enter this release. Demographics, free-text notes, diagnosis inference, and live credentialed MIMIC-IV access are not required.
07 / Historical Load
Undefined is the result
null, not zero08 / Reading discipline
What this release refuses to turn into a claim
Dates
Sequence and elapsed intervals are usable; displayed years are not real calendar dates.
Prescriptions
An order interval is not administration, adherence, or treatment response.
Lab frame
Four latest values are an explicit observation window, never complete patient state.
Similarity
A distance under one metric creates no identity or clinical-equivalence relation.
Clinical meaning
Temporal order and record counts supply no diagnosis, cause, prediction, or recommendation.