This paper presents a new method for coronary artery segmentation in X-ray angiograms based on deep learning and a patch-based training. The blood vessel segmentation is performed using the U-Net convolutional neural network, which has been trained using patches extracted from the original angiograms instead of using complete images. The publicly available database of coronary angiograms DCA1 containing 130 angiograms with their respective ground-truth has been used to generate the training patterns and subsequently to evaluate and compare the segmentation performance of the proposed method. The hyper-parameter configuration used for training the U-Net parameters has been selected from 90 possible combinations according to five binary classification metrics. Each combination involving the selection of a patch size, weight assigned to the blood vessel class, and learning rate used by the optimization method, has been used in order to train the U-Net parameters with patterns extracted from a set of 100 images. The segmentation performance of the proposed method is compared with five specialized blood vessel segmentation methods from the state of the art using a test set of 30 images, achieving the highest accuracy (0.977) and Dice similarity coefficient (0.779). Moreover, the experimental results have also shown that the proposed method is suitable to be integrated into a computer-aided system to support decision making in medical practice.
This paper presents a novel method for the automatic design of convolutional gray-level templates for detecting coronary arteries in X-ray angiograms. The proposed method uses the metaheuristic of iterated local search (ILS) to address the high-dimensional problem (O(256n)) involved in the design of convolutional templates. This automatically generated template is convolved in the spatial domain at different orientations to form a directional filter bank in order to detect coronary arteries at different angular resolutions. The vessel detection results are compared with those obtained by four state-of-the-art vessel enhancement methods in terms of the area (Az) under the receiver operating characteristic (ROC) curve. The proposed method achieved the highest detection results with Az = 0.9405 using a training set of 50 angiograms. Moreover, the convolutional gray-level template obtained from the training step, it was directly evaluated with an independent test set of 50 X-ray angiograms obtaining an Az = 0.9565, which is the highest performance according to the comparative analysis. In addition to the experimental results, the use of metaheuristics for designing convolutional gray-level templates obtains suitable results to be considered in systems that perform computer-aided diagnosis, and it also represents an encouraging area for future research.
This paper presents a novel method for the automatic design of binary descriptors for the detection of coronary arteries in X-ray angiograms. The method is divided into two different steps for detection and segmentation. In the step of automatic vessel detection, the metaheuristic of iterated local search (ILS) is used for the design of optimal binary descriptors for detecting vessel-like structures by using the top-hat transform in the spatial image domain. The detection results are compared with those obtained by five state-of-the-art vessel enhancement methods. The proposed method obtained the highest detection results in terms of the area (Az ) under the ROC curve (Az = 0.9635) using a training set of 50 angiograms, and Az = 0.9544 with an independent test set of 50 X-ray images. In the segmentation step, the inter-class variance thresholding method was applied to classify vessel and nonvessel pixels from the top-hat filter response obtained from the binary descriptor. According to the experimental results, the vessel detection by using an automatically generated binary descriptor can be highly suitable for computer-aided diagnosis.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.