Deep Learning-Based Assessment of Coronary Artery Morphologies for Predicting Responsiveness to Intravascular Lithotripsy
Authors' conclusion
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.
Study context
- Population
- Patients with varying degrees of coronary artery calcification treated with intravascular lithotripsy (IVL)
- Participants
- 134 patients
- Comparison
- Prediction performance for poststenting luminal expansion incorporating comprehensive morphological features vs conventional geometric dimensions
- Outcome
- Prediction performance for luminal expansion following IVL, frame level
Methodological detailsartificial intelligence-enhanced image analysis, optical coherence tomography before IVL and after stent deployment, linear regression
- Artificial intelligence-enhanced image analysis
- Optical coherence tomography before IVL and after stent deployment
- Linear regression
Main finding
As reported in the abstract
| Group | Adjusted R2 | Root mean square error |
|---|---|---|
| Comprehensive morphological features | 0.717 ± 0.008 | 0.095 ± 0.004 |
| Conventional geometric dimensions | 0.581 ± 0.016 | 0.133 ± 0.007 |
Adjusted R2
Comprehensive morphological features0.717 ± 0.008
Conventional geometric dimensions0.581 ± 0.016
Root mean square error
Comprehensive morphological features0.095 ± 0.004
Conventional geometric dimensions0.133 ± 0.007
Further findings
| Outcome | Comprehensive morphological features | Conventional geometric dimensions |
|---|---|---|
| Prediction performance for luminal expansion following IVL, lesion level | ||
| Adjusted R2 | 0.721 ± 0.010 | 0.582 ± 0.018 |
| Root mean square error | 0.093 ± 0.009 | 0.135 ± 0.021 |
- Prediction performance for luminal expansion following IVL, lesion level: adjusted R2 P < 0.001
- Prediction performance for luminal expansion following IVL, lesion level: root mean square error P < 0.001
Morphological predictors of poststenting luminal expansion
- Lumen and calcium circularity, eccentricity of vessel wall, and balance of calcium to lumen area identified as significant predictors
- P < 0.001
Procedural parameters and poststenting luminal expansion
- Maximum stent diameter and IVL catheter size associated with poststenting luminal expansion
- P < 0.001
Original publication
Deep Learning-Based Assessment of Coronary Artery Morphologies for Predicting Responsiveness to Intravascular Lithotripsy. JACC Adv. Published 6 August 2026. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →