ABC stenosis morphology classification and outcome of coronary angioplasty: reassessment with computing techniques

Area of Science:

• Cardiovascular research
• Interventional cardiology
• Medical imaging analysis

Background:

• The ACC/AHA stenosis morphology classification (MC) is used to assess coronary lesions before percutaneous coronary angioplasty (PTCA).
• Its individual predictive value for PTCA outcomes has not been fully evaluated using modern computational methods.

Purpose of the Study:

• To evaluate the individual predictive value of MC for procedural success (PS) and major adverse cardiac events (MACE) after PTCA.
• To compare the predictive performance of conventional logistic regression with machine learning techniques.

Main Methods:

• Morphology classification (MC) was assigned to 325 coronary lesions by two independent observers.
• Predictive values for PS and MACE were analyzed using logistic regression and machine learning models.
• The study included 250 patients undergoing PTCA.

Main Results:

• Procedural success (PS) decreased and major adverse cardiac events (MACE) increased from MC type A to C.
• Conventional logistic regression showed a high error rate (42%) and no single MC factor was predictive.
• Machine learning models achieved a 10% predictive error for MC alone, reduced to 2% with additional parameters.
• MC parameters were highly ranked for predicting PS, while medical history variables were more impactful for MACE.

Conclusions:

• Stenosis morphology classification (MC) alone is insufficient for predicting PTCA procedural success (PS) or major adverse cardiac events (MACE).
• Integrating all MC parameters with lesion-specific and patient history variables significantly enhances predictive accuracy.
• This approach can aid risk stratification in the catheterization lab and improve interventional cardiology classification systems.