Super Learner Algorithm for Carotid Artery Disease Diagnosis: A Machine Learning Approach Leveraging Craniocervical...

Area of Science:

• Medical Imaging
• Machine Learning in Healthcare
• Vascular Diagnostics

Background:

• Carotid artery diseases like stenosis, aneurysm, and dissection pose significant health risks.
• Accurate and timely diagnosis is crucial for effective patient management.
• Current diagnostic methods may have limitations in sensitivity and specificity.

Purpose of the Study:

• To develop and validate a machine learning model for diagnosing carotid artery diseases using craniocervical CTA.
• To evaluate the performance of a super learner model integrating multiple ML algorithms.
• To improve diagnostic accuracy and robustness for conditions like aneurysm and dissection.

Main Methods:

• Utilized a curated dataset of 122 craniocervical CTA patient cases.
• Developed a super learner model combining adaptive boosting, gradient boosting, and random forests.
• Applied techniques including k-fold cross-validation, bootstrapping, data augmentation, and SMOTE for enhanced robustness and performance on minority classes.

Main Results:

• The super learner model achieved an overall accuracy of 90% in diagnosing carotid artery diseases.
• Significant performance improvements were observed for minority classes such as aneurysm and dissection.
• The proposed ML approach demonstrated superior accuracy and robustness compared to state-of-the-art methods.

Conclusions:

• Machine learning, particularly the super learner model, shows great promise for accurate carotid artery disease diagnosis from CTA.
• Blood vessel structural analysis is a key factor in ML-based diagnostic accuracy.
• This research provides a foundation for future advancements in AI-driven medical diagnostics.