Related Experiment Videos
Beneficial application of quantitative coronary angiography (edge detection algorithm) in analysis of dissected
1Division of Cardiology, Toyohashi Heart Center, Toyohashi, Japan.
Insights
Quantitative coronary angiography using an edge detection algorithm effectively analyzes coronary dissections post-angioplasty. Automated analysis alone is sufficient for predicting long-term patency in NHLBI types B-C lesions.
Area of Science:
- Cardiovascular Imaging
- Interventional Cardiology
- Medical Image Analysis
Background:
- Coronary artery dissection is a complication of balloon angioplasty.
- Accurate assessment of dissection severity is crucial for predicting outcomes.
- Quantitative coronary angiography (QCA) with edge detection algorithms offers automated analysis.
Purpose of the Study:
- To evaluate the accuracy of an edge detection algorithm in quantitative coronary angiography for analyzing coronary dissection lesions after balloon angioplasty.
- To compare the performance of automated analysis versus manual editing in predicting long-term patency.
Main Methods:
- Retrospective analysis of 66 coronary dissection lesions (NHLBI types B-C) in 60 patients.
- Application of an edge detection algorithm for automated lumen border delineation.
- Comparison of automated analysis with and without manual editing.
Main Results:
- The edge detection algorithm delineated the true lumen in 48.5% of lesions and included dissection flaps in 51.5% with manual editing.
- Automated analysis showed significant correlation between minimal lumen diameter post-procedure and at 5.3-month follow-up in both groups (r=0.554 and r=0.613).
- Manual editing reduced the correlation coefficient (r=0.240), indicating less predictive power for long-term patency.
Conclusions:
- The edge detection algorithm in quantitative coronary angiography is a reliable tool for analyzing coronary dissections.
- Automated analysis alone is sufficient and potentially more accurate for predicting long-term patency in NHLBI types B and C dissected lesions.
- Manual editing of edge detection measurements may not improve, and could potentially hinder, the prediction of long-term outcomes.
Abstract:
The present study evaluated the application of quantitative coronary angiography (edge detection algorithm) for the analysis of coronary dissection lesions after balloon angioplasty. Acute and late results were obtained by the edge detection algorithm in 60 patients with 66 dissected lesions (NHLBI types B-C). The edge detection algorithm delineated the border of the true lumen in 32 lesions (group with automated analysis alone, 48.5%) and included the dissection cap in the analysis in 34 lesions in which manual editing was adjuncted (group with manual editing, 51.5%). In both groups, the minimal lumen diameter after balloon angioplasty obtained by initial automated analysis was correlated to that obtained at the 5.3-month follow-up similarly (r=0.554, p=0.0010 for the group with automated analysis alone and r=0.613, p=0.0001 after automated analysis for the group with manual editing). However, additional manual editing reduced the correlation coefficient (r=0.240, p=0.1707) in the latter group. Thus, in terms of predicting long-term patency, it is reasonable to let the edge detection algorithm decide the measurements in types B and C dissected lesions.