Comprehensive Summary
Saleh and Khatri report the first clinical use of an artificial intelligence, based spatiotemporal enhancement processing (STEP) software (AngioWave Imaging) to support procedural planning for chronic total occlusion (CTO) percutaneous coronary intervention (PCI). A 59-year-old man with non–ST-segment elevation myocardial infarction underwent culprit LAD PCI and was referred for circumflex CTO PCI, where a standard angiography could not clearly define any collateral pathways. The angiogram was processed offline with STEP, which applies a neural network trained on 120 annotated angiographic examples to enhance vessel signal and collateral visualization. Using this AI-guided strategy, the operators successfully crossed the CTO into the distal true lumen, deployed a 2.5 × 48 mm drug-eluting stent, and achieved final TIMI 3 flow with a fluoroscopy time of 24.4 minutes, total procedure time 80 minutes, and 160 mL contrast use.
Outcomes and Implications
For interventional cardiologists, this case suggests that AI-based STEP enhancement can uncover subtle antegrade channels and clinically relevant collaterals that are missed on standard angiography, directly altering CTO strategy selection toward safer, more efficient wiring options. At the bedside and in the cath lab, similar tools could reduce unnecessary retrograde attempts, shorten procedure and fluoroscopy times, and potentially decrease contrast exposure, although these benefits require confirmation in larger, prospective CTO PCI cohorts beyond this single-center case.