Uncertainty in Automated Stenosis Quantification Using Multiview X-ray Coronary Angiography Videos

The visual interpretation of X-ray coronary angiography, the primary imaging modality for coronary stenosis evaluation, is a difficult ta sk an d re qu ires ex pe rience an d ex pe rt kn ow ledge. Au to mating st en osis assessment can improve confidence i n s t enosis i d entification an d se ve rity es ti mation, fa ci litating de ci sions re ga rding revascularization strategies. However, existing methods are predominantly limited to static images or single-view videos, which increases the risk of missing crucial information due to the complex structure of the coronary tree and movement of the heart. We propose a five-step w orkflow fo r au tomated st enosis de tection, localization and severity estimation in X-ray angiography videos. For evaluation at the patient-level, multiple videos per patient, captured from different v i ews, w ere c o nsidered. T h e w orkflow ac hieved an ov erall se ns itivity of 58.98% and specificity o f 8 4 .15% f o r s t enosis p r ediction p e r c o ronary s e gment. S e nsitivity a n d s p ecificity fo r stenosis severity classification w e re 6 2 .75% a n d 5 9 .72%, r e spectively. T o a s sess t h e i m pact o f m u lti-view a n alysis, we compared severity estimation performance for stenoses detected in single- and multi-view projections, demonstrating that only one view is associated with the highest uncertainty. Our findings e ncourage f urther refinement and development of the workflow a nd h ighlight t he i mportance o f m ulti-view c onsideration f or a ccurate stenosis evaluation.