Sign in
EUS-PDA-QE-2026Active studySynthetic mock data

Quantitative EUS Endpoints

AI-derived tumor probability as a longitudinal endpoint for pancreatic cancer trials

Subjects enrolled

10

5 treatment · 5 control

EUS scans analyzed

30

3 per subject · baseline + wk 8 + wk 16

Treatment Δ

-45 pp

mean change from baseline

Control Δ

+17 pp

mean change from baseline

AI tumor probability over time

Mean ± IQR by arm with individual patient traces · scissors divergence at week 16

Waterfall — Δ tumor probability from baseline

Per-subject change in AI-derived tumor probability (percentage points). Bars below 0 indicate tumor probability reduction.

Treatment (5)Control (5)
-60-50-40-30-20-100+10+20+30Response -20ppProgression +10ppPatricia Nguyen (EUS-PDA-005) — -53pp — Treatment-53005Maria Garcia (EUS-PDA-001) — -50pp — Treatment-50001Linda Okafor (EUS-PDA-003) — -46pp — Treatment-46003Robert Chen (EUS-PDA-002) — -44pp — Treatment-44002James Wilson (EUS-PDA-004) — -32pp — Treatment-32004David Kim (EUS-PDA-008) — +14pp — Control+14008Susan Martinez (EUS-PDA-007) — +16pp — Control+16007Richard Taylor (EUS-PDA-010) — +16pp — Control+16010Nancy Patel (EUS-PDA-009) — +18pp — Control+18009Thomas Brown (EUS-PDA-006) — +19pp — Control+19006pp change
10 subjects, sorted by delta ascending

Subjects

EUS-PDA-001Treatment

Maria Garcia

28%-50 pp
EUS-PDA-002Treatment

Robert Chen

38%-44 pp
EUS-PDA-003Treatment

Linda Okafor

22%-46 pp
EUS-PDA-004Treatment

James Wilson

42%-32 pp
EUS-PDA-005Treatment

Patricia Nguyen

32%-53 pp
EUS-PDA-006Control

Thomas Brown

84%+19 pp
EUS-PDA-007Control

Susan Martinez

88%+16 pp
EUS-PDA-008Control

David Kim

92%+14 pp
EUS-PDA-009Control

Nancy Patel

86%+18 pp
EUS-PDA-010Control

Richard Taylor

78%+16 pp