CLARIS
Clinical Investigation Workstation
DR
Demo case selection

Pick one prepared HCC-TACE-Seg case. The workstation updates from backend-generated static API artifacts.

Data boundaryloading
Case Overview

HCC_015

Review RequiredMeasurements available

This page is the readable summary of the selected CLARIS case. Use the navigation pages for the full patient contract, imaging workspace, investigation timeline, evidence trace and report.

Run completed
Checkpoint 9
Assessment
Review Required
Evidence
LIMITED
Supported claims
100%
15/15 claims
Missing info
14
Conflicts
1
Evidence sources
7
What is known now
Current imaging measurements
Available
Lesion count
1
Largest lesion volume
62.3191 cm3
derived from SEG
Approximate diameter
55.117 mm
derived measurement
AFP (latest)
3.6 ng/mL
observed
Treatment context
TACE observed
source
Measurements require clinical review. Bounding-box diameter is not a validated RECIST measurement.
Assessment in plain language
Completed With Review Required.

CLARIS synthesized structured case investigation and controlled evidence. Treatment context is available. Current mask-derived imaging measurements are available for review. AFP is available as contextual clinical data. Longitudinal treatment-response assessment remains insufficient because compatible previous/current measurements are unavailable. Clinical interpretation is limited by unavailable liver-function data. Structured conflicts or provenance gaps require clinician review. Decision status: completed_with_review_required.

Supported context
  • CLARIS completed a case investigation for HCC_015 with status completed_with_review_flags.
  • AFP is available as an observed source value: 3.6 ng/mL. Structured clinical context is available from the normalized PatientCase with 52 observed clinical records. Liver-function assessment is incomplete because required clinical fields are unavailable: bilirubin, albumin, INR, platelets, complete_liver_function_panel.
  • Current mask-derived imaging measurements are available: lesion count 1, largest lesion volume 62.3191 cm3, approximate bounding-box diameter 55.117 mm.
  • Treatment context is available from source records. TACE treatment context is observed. Source response observations are available and preserved as source fields; CLARIS has not recalculated formal response criteria. Formal longitudinal treatment response cannot be established by CLARIS because previous/current compatible imaging measurements are unavailable.
Patient Timeline
6 events
Open a focused page
Needs review
Missing information14
Review flags1
Conflicts1
Source response fields exist, but CLARIS did not complete its own previous/current quantitative imaging comparison.
Provenance
Source response fields are available, but CLARIS has not completed a compatible previous/current imaging comparison.
Provenance Gap
Recent investigation activity
Clinical ContextClinical Context Agent14:27:11
Available Evidence Count: 52 · Missing Information Count: 10 · Treatment Context Count: 2
Imaging InvestigationImaging Investigation Agent14:28:18
Previous Study Count: 1 · Current Request Status: metadata_available_measurements_unavailable · Comparison Requested: True
Imaging ExecutionImaging Execution Layer14:29:25
Current Status: completed · Comparison Status: insufficient_data · Executed Request Count: 1
Treatment Response ContextTreatment Response Context Agent14:30:32
Source Response Count: 21 · Review Flag Count: 1 · Unsupported Criteria Count: 3
Key limitations
  • Checkpoint 7 orchestrates available CLARIS capabilities; it does not synthesize a final clinical recommendation.
  • Routing is deterministic for auditability and will be compared with LLM-assisted routing later.
  • Checkpoint 8 retrieves only from the controlled CLARIS evidence corpus.
  • Evidence records are curated source notes with provenance, not full redistributed guideline documents.
  • The Evidence Investigation Agent does not generate treatment recommendations.
  • Evidence support means relevant controlled-corpus support was found; it is not clinical validation.