Towards a Deep-Learning Approach for Prediction of Fractional Flow Reserve from Optical Coherence Tomography

Abstract

Cardiovascular disease (CVD) is the number one cause of death worldwide, and coronary artery disease (CAD) is the most prevalent CVD, accounting for 42% of these deaths. In view of the limitations of the anatomical evaluation of CAD, Fractional Flow Reserve (FFR) has been introduced as a functional diagnostic index. Herein, we evaluate the feasibility of using deep neural networks (DNN) in an ensemble approach to predict the invasively measured FFR from raw anatomical information that is extracted from optical coherence tomography (OCT). We evaluate the performance of various DNN architectures under different formulations: regression, classification—standard, and few-shot learning (FSL) on a dataset containing 102 intermediate lesions from 80 patients. The FSL approach that is based on a convolutional neural network leads to slightly better results compared to the standard classification: the per-lesion accuracy, sensitivity, and specificity were 77.5%, 72.9%, and 81.5%, respectively. However, since the 95% confidence intervals overlap, the differences are statistically not significant. The main findings of this study can be summarized as follows: (1) Deep-learning (DL)-based FFR prediction from reduced-order raw anatomical data is feasible in intermediate coronary artery lesions; (2) DL-based FFR prediction provides superior diagnostic performance compared to baseline approaches that are based on minimal lumen diameter and percentage diameter stenosis; and (3) the FFR prediction performance increases quasi-linearly with the dataset size, indicating that a larger train dataset will likely lead to superior diagnostic performance.

Artificial Intelligence in Estimating Fractional Flow Reserve: A Systematic Literature Review

“…In recent years, end-to-end frameworks have been introduced in the eld of DL, and the bene ts of using them in health have been investigated [58,59]. The present study shows that several studies used this framework to estimate FFR [27,29,32,34,35,37,38,40,41,43]. Due to the need for the end-to-end framework for a large amount of data and the lack of data in these studies, the over tting problem should also be considered [34], for which we need many data.…”

“…Firstly, they always return the same qualitative results from a speci c input; secondly, like humans, there is no variance due to fatigue [55]. This study also shows that in several studies, image segmentation steps and feature extraction using DL methods have been done [27,29,32,34,35,37,38,40,41,43]. In addition, in some studies, the parameters in the images were extracted using manual methods and commercial software [25,28,30,42,46] [39].…”

“…Several studies show that the accuracy of predicting the FFR can be different according to the type of branches [54]. This study also shows that the results obtained separately for each vessel and each segment (proximal, mid, and distal) are different [25,32,41,48]. Therefore, the separation of the type of vessels and the separation of each vessel according to the type of segment to determine the accuracy of the FFR estimation model is essential.…”

“…This study demonstrated that various imaging tools have been used, including CCTV [24][25][26][27][28][29][30][31][32][33][34][35][36][37][38][39][40], OCT [41][42][43], XCA [44], and IVUS [45]. Several studies used a combination of CCTV with OCT [46], IVUS [47], and XCA [48], and a study combined IVUS and XCA [49] to estimate the FFR.…”

Deep Neural Networks in Medical Imaging: Privacy Preservation, Image Generation and Applications

“…The main objective in the third paper is to evaluate the feasibility of using neural networks in predicting invasively measured FFR from the radius of the coronary lumen that is extracted along the centerline of the coronary artery from OCT images [15]. Three different approaches were used for solving this task: a regression, a classification and an FSL (few shot learning) approach, where the task was formulated also as a classification problem.…”