Data source and utilization of artificial intelligence technologies in vascular surgery-a scoping review

Area of Science:

• Medical Informatics
• Artificial Intelligence in Surgery
• Vascular Surgery Research

Background:

• Artificial Intelligence (AI) applications are expanding in healthcare, including vascular surgery.
• Natural Language Processing (NLP) can extract valuable information from unstructured medical text.
• AI can enhance risk prediction, support decision-making, and improve clinical trial recruitment.

Purpose of the Study:

• To review data sources for AI algorithms in vascular surgery.
• To identify AI applications within vascular surgery.
• To determine target journals for AI in vascular surgery research.

Main Methods:

• Scoping review of peer-reviewed articles from January 1996 to March 2023.
• Analysis of 342 eligible articles focusing on AI, machine learning (ML), and deep learning (DL).
• Categorization of AI applications by surgical area, data source, and function.

Main Results:

• Machine learning (ML) and deep learning (DL) were prevalent AI techniques.
• Aorta, carotid disease, and peripheral arterial disease were key research areas.
• AI algorithms were primarily used for predictive modeling, image segmentation, and diagnosis support.
• Image-based data was most common, followed by medical records and clinical parameters.

Conclusions:

• AI adoption in vascular surgery is growing, with diverse data sources and applications.
• Natural Language Processing (NLP) is underutilized in vascular surgery research despite available text data.
• Optimizing data source utilization is crucial for advancing AI in vascular surgery.