Pre-trained Artificial Intelligence Models in the Prediction and Classification of Atherosclerotic Cardiovascular Disease

Atherosclerotic cardiovascular disease (ASCVD) is one of the leading causes of global morbidity and mortality. The current study provides a systematic review of the use of artificial intelligence (AI) technologies applied to the prediction and management of ASCVD. Traditional risk assessment approaches have their restrictions, leading to a growing preference for AI and machine learning techniques in risk assessment. First, this study tackles the complex pathophysiology of ASCVD and the problems associated with the current diagnosis, followed by an in-depth analysis of the wide variety of AI models that can be applied to electronic health records, medical imaging data, and other biomarkers. Special attention will be paid toward the potential of natural language processing models like bidirectional encoder representations from transformers in predicting risk from textual clinical data, and the overwhelming success of convolutional neural networks such as residual neural network and visual geometry group in plaque-based analysis through imaging modalities. Although the research results show that these models have a lotto offer in the clinical world, the authors also describe some serious disadvantages: data bias, interpretability of the model, and computational needs. It highlights, in particular, the need for multicenter validation studies as well as developing explainable AI techniques. Overall, AI-based approaches may pave the way for a new paradigm in ASCVD management. Nevertheless, deploying these technologies in everyday clinical practice will require overcoming technical, ethical, and regulatory challenges. As such, interdisciplinary collaboration and thorough clinical validation studies are essential for fulfilling the promise of these novel strategies to enhance patient outcomes.

Dergi

EURASIAN JOURNAL OF MEDICINE

Yıl

2025

Cilt

57

Sayı

3

eISSN

1308-8742

DOI

10.5152/eurasianjmed.2025.25937

WoS UID

WOS:001685819500004