Artificial intelligence in estimating instantaneous wave-free ratio: a systematic literature review of techniques

Fractional flow reserve (FFR) is an essential tool for evaluating coronary artery disease and directing percutaneous coronary intervention (PCI). The instantaneous wave-free ratio (iFR) has been validated as a non-hyperemic alternative with less procedural complexity and adverse effects, as no pharmacological induction of hyperemia is required. Nevertheless, iFR invasiveness limits the popularity of the technique in clinical practice. However, recent AI breakthroughs have led to improvements in the diagnostic accuracy of non-invasive iFR estimation via different imaging modalities such as X-ray coronary angiography (XCA) and coronary computed tomography angiography (CCTA). A systematic search was conducted in the Web of Science, PubMed, ScienceDirect, and Scopus databases without any date restriction. Only studies that resulted in the development of AI-based methods for the estimation of iFR were considered. Five studies met the inclusion criteria and used AI to estimate iFR from CCTA and XCA image data. The diagnostic accuracy reported varied from 58% to 90.2%, while sensitivity was between 37% and 87.2%, and specificity between 50% and 97.8%. Positive predictive value (PPV) and negative predictive value (NPV) ranged from 34% to 79% and 77% to 97.5%, respectively. The value of the receiver operating characteristic (ROC) curve ranged from 0.89 to 0.98. The QUADAS-2 tool was used to evaluate the quality of the study. AI models reported a promising improvement in the assessment of coronary artery disease based on accurate non-invasive methodologies. However, further research is needed to establish standardized practices and ensure the accessibility and applicability of these tools.