Artificial Intelligence-Based Fully Automated Quantitative Coronary Angiography vs Optical Coherence Tomography-Guided PCI: The FLASH Trial

Abstract

Background: Recently developed artificial intelligence–based coronary angiography (AI-QCA, fully automated) provides real-time, objective, and reproducible quantitative analysis of coronary angiography without requiring additional time or labor.

Objectives: This study aimed to evaluate the efficacy of AI-QCA–assisted percutaneous coronary intervention (PCI) compared to optical coherence tomography (OCT)-guided PCI in terms of post-PCI results.

Methods: This trial enrolled 400 patients with significant coronary artery disease undergoing PCI from 13 participating centers in South Korea. Patients were randomized in a 1:1 ratio to either AI-QCA–assisted or OCT-guided PCI. The primary endpoint was the post-PCI minimal stent area (MSA) assessed by OCT. The noninferiority of AI-QCA–assisted PCI to OCT-guided PCI regarding the post-PCI MSA was tested with a noninferiority margin of 0.8 mm2.

Results: A total of 395 patients (199 in the AI-QCA group and 196 in the OCT group) were included in the primary endpoint analysis. The post-PCI MSA was 6.3 ± 2.2 mm2 in the AI-QCA group and 6.2 ± 2.2 mm2 in the OCT group (difference, −0.16; 95% CI: −0.59 to 0.28; P for noninferiority < 0.001). Other OCT-defined endpoints, such as stent underexpansion (50.8% [101/199] vs 54.6% [107/196]; P = 0.48), dissection (15.6% [31/199] vs 12.8% [25/196]; P = 0.42), and untreated reference segment disease (15.1% [30/199] vs 13.3% [26/196]; P = 0.61), were not significantly different between groups, except for a higher incidence of stent malapposition in the AI-QCA group (13.6% [27/199] vs 5.6% [11/196]; P = 0.007).

Conclusions: This study demonstrated the noninferiority of AI-QCA–assisted PCI to OCT-guided PCI in achieving MSA with comparable OCT-defined endpoints. (Fully Automated Quantitative Coronary Angiography Versus Optical Coherence Tomography Guidance for Coronary Stent Implantation