Deep Learning-Based Assessment of Coronary Artery Morphologies for Predicting Responsiveness to Intravascular Lithotripsy.

Authors

Yiqing Liu, Farhad R Nezami, Elazer R Edelman

Published in

JACC. Advances. Volume 5. Issue 9. Pages 103155. Aug 06, 2026. Epub Aug 06, 2026.

Abstract

Although intravascular lithotripsy (IVL) has proven effective in facilitating luminal expansion through calcification modification, its effectiveness in luminal expansion across different coronary morphologies remains poorly understood.
The authors sought to identify predictive lumen and calcium morphologies for luminal expansion following IVL through artificial intelligence-enhanced image analysis.
A total of 134 patients with varying degrees of coronary artery calcification were included. Optical coherence tomography before IVL and after stent deployment were analyzed. Using a deep-learning approach, the predictive value of pre-IVL lumen and calcium morphologies that capture structural complexity beyond basic geometric dimensions for predicting poststenting luminal expansion was comprehensively evaluated. The impact of morphological predictors was analyzed using linear regression.
The prediction performance for luminal expansion following IVL was significantly improved above conventional geometric dimensions by incorporating comprehensive morphological features at both frame (adjusted R² = 0.717 ± 0.008 vs 0.581 ± 0.016, P < 0.001; root mean square error = 0.095 ± 0.004 vs 0.133 ± 0.007, P < 0.001) and lesion levels (adjusted R² = 0.721 ± 0.010 vs 0.582 ± 0.018, P < 0.001; root mean square error = 0.093 ± 0.009 vs 0.135 ± 0.021, P < 0.001). Morphological descriptors that capture intricate anatomical structural variations, including lumen and calcium circularity, eccentricity of vessel wall, and balance of calcium to lumen area, were identified as significant predictors for poststenting luminal expansion (P < 0.001). Procedural parameters including maximum stent diameter and IVL catheter size were associated with poststenting luminal expansion (P < 0.001).
A deep learning-based approach enhances the predictability of IVL effectiveness in luminal expansion by capturing preprocedural morphological complexities beyond conventional dimensions, supporting personalized procedural planning. Further prospective validation is warranted to confirm its generalizability and clinical impact.

PMID:
42561486
Bibliographic data and abstract were imported from PubMed on 07 Aug 2026.

Read full publication at:
Please sign in to see all details.

Stats

1-terrible, 9-excellent. How would you rate this publication? Sign in in to submit your rating.

Post a comment

You need to be signed in to post comments. You can sign in here.

Comments

There are no comments yet.