Automatic Feature Selection for Stenosis Detection in X-ray Coronary Angiograms

Abstract

The automatic detection of coronary stenosis is a very important task in computer aided diagnosis systems in the cardiology area. The main contribution of this paper is the identification of a suitable subset of 20 features that allows for the classification of stenosis cases in X-ray coronary images with a high performance overcoming different state-of-the-art classification techniques including deep learning strategies. The automatic feature selection stage was driven by the Univariate Marginal Distribution Algorithm and carried out by statistical comparison between five metaheuristics in order to explore the search space, which is O(249) computational complexity. Moreover, the proposed method is compared with six state-of-the-art classification methods, probing its effectiveness in terms of the Accuracy and Jaccard Index evaluation metrics. All the experiments were performed using two X-ray image databases of coronary angiograms. The first database contains 500 instances and the second one 250 images. In the experimental results, the proposed method achieved an Accuracy rate of 0.89 and 0.88 and Jaccard Index of 0.80 and 0.79, respectively. Finally, the average computational time of the proposed method to classify stenosis cases was ≈0.02 s, which made it highly suitable to be used in clinical practice.

Non-invasive fractional flow reserve estimation using deep learning on intermediate left anterior descending coronary artery lesion angiography images

“…It has many applications for automatically extracting rules and features from various data types. CNNs are extensively used for image processing 36 and classifying medical images 37 . They are used to segment coronary vessels 9 and classify and identify stenosis in vessels 36 , 38 using angiography images.…”

“…CNNs are extensively used for image processing 36 and classifying medical images 37 . They are used to segment coronary vessels 9 and classify and identify stenosis in vessels 36 , 38 using angiography images. Using pre-trained CNN models to increase accuracy and effectively reduce training time is a common approach in artificial neural networks.…”

“…Using pre-trained CNN models to increase accuracy and effectively reduce training time is a common approach in artificial neural networks. This method is referred to as transfer learning 36 .…”

Coronary Artery Disease Classification With Different Lesion Degree Ranges Based on Deep Learning

“…Some works had been dedicated to ICA feature extraction and Principal Component Analysis (PCA) to use Machine Learning methods as classifiers [11]- [13]. Gil-Rios et al [12] used the Univariate Marginal Distribution Algorithm and statistical comparison between five metaheuristics to explore the search space in order to develop an automatic feature selection of ICA images. As a result of the PCA study, a subset of 20 features was established to correctly classify ICA images, with an accuracy of 89%.…”

The Application of Deep Learning for the Segmentation and Classification of Coronary Arteries

“…Gil-Rios, Miguel-Angel, and colleagues proposed the use of a Support Vector Machine (SVM) for the detection of coronary stenosis from the Antczak and Liberadzki image dataset as well as the dataset of the Mexican Social Security Institute. The model performance yielded remarkable results [ 31 ]. Ovalle-Magallanes, Emmanuel, and colleagues presented a method to automatically detect coronary artery stenosis based on X-ray coronary angiograms.…”