Canonical text (the exact bytes fingerprinted)

TRUVACE RECORD VERSION
record: TRV-2026-1833
version: 1
kind: certified
reason: Certified into the record
timestamp: 2026-10-18T18:13:03.686044Z
status: published
lens: g_space
sector: health
headline: Sex Differences in Detection of Acute Coronary Syndrome Culprit Lesion Using Conventional and AI-Enhanced Coronary CT Angiography
dek: Aims Acute coronary syndrome (ACS) exhibits marked sex-related differences in clinical presentation and underlying pathophysiology. However, whether these differences translate into variations in culprit lesion prediction remains unclear. We investigated the sex-specific performance of conventional coronary computed tomography angiography (CCTA) and artificial intelligence-enabled quantitative coronary plaque and hemodynamic assessment (AI-QCPHA) for ACS culprit detection. Methods and results In this sub-study o…
gain_title: Adding AI-derived plaque and hemodynamic features raised discrimination of ACS culprit lesions over conventional coronary CT in both women and men, using the same best features for each sex.
problem_title: (none)
trace_subject: (none)
gain_reading: Adding AI-derived plaque and hemodynamic features raised discrimination of ACS culprit lesions over conventional coronary CT in both women and men, using the same best features for each sex.
gain_evidence: integration of these features improved discriminatory performance beyond conventional CCTA (women AUC 0.84 vs. 0.78, p = 0.002; men 0.82 vs. 0.75 | AI-QCPHA identified the same best features for ACS culprit prediction in both sexes
problem_reading: (none)
problem_evidence: (none)
quick_read: Researchers analyzed 351 patients with acute coronary syndrome, 90 women and 261 men, who had undergone coronary CT angiography between one month and three years before their event. Lesions were later classified as culprit or non-culprit using invasive angiography at the time of the ACS, which let the team test how well scans predicted the lesion that eventually caused the event.

Conventional CT criteria for obstructive stenosis performed similarly in both sexes, while high-risk plaque criteria differed by sex. An AI-based quantitative plaque and hemodynamic tool improved prediction in both groups. This matters because ACS presents differently in women and men, and the findings suggest one AI approach may work across both. The abstract is truncated and the cohort was predominantly male, so some details remain unclear.
limitation: The cohort included far fewer women (90) than men (261), and the conclusion is framed as supporting a potential role rather than demonstrating clinical benefit.
tag: Evidence-backed gain
key_points: The sub-study drew on the EMERALD II trial (NCT03591328) and covered 351 patients with 2,451 lesions. | Culprit and non-culprit lesions were defined by invasive coronary angiography at the time of ACS. | Conventional CCTA performed similarly across sexes for obstructive stenosis, whereas high-risk plaque criteria differed by sex. | Feature selection across five feature clusters identified the same best AI predictors in women and men.
rundown: The study compared conventional coronary CT angiography criteria, namely obstructive stenosis (CAD-RADS 3 or higher) and high-risk plaque features, with an AI-enabled quantitative plaque and hemodynamic assessment. Performance was compared between women and men.

Information gain-based feature selection was used to find sex-specific best predictors across five feature clusters, and their added predictive value was evaluated with area under the curve. In women, the AUC rose from 0.78 to 0.84 (p = 0.002), and in men from 0.75 to 0.82.

The authors conclude that the AI approach strengthens culprit lesion prediction in obstructive ACS in both sexes, supporting a potential role for better culprit-lesion discrimination.
sources:
- peer_reviewed | European Heart Journal - Cardiovascular Imaging | https://doi.org/10.1093/ehjci/jeag270 | 2026-10-06
prev: 0000000000000000000000000000000000000000000000000000000000000000

sha256

e0937695caf33d6bad93389dec4042ddd0f485cfc1c30cc45c507d337569bc7e

previous

0000000000000000000000000000000000000000000000000000000000000000