Automatic Coronary Calcium Scoring in Cardiac CT Angiography Using Convolutional Neural Networks
The amount of coronary artery calcification (CAC) is a strong and independent predictor of cardiovascular events. Non-contrast enhanced cardiac CT is considered a reference for quantification of CAC. Recently, it has been shown that CAC may be quantified
- PDF / 1,657,431 Bytes
- 8 Pages / 439.363 x 666.131 pts Page_size
- 40 Downloads / 225 Views
Image Sciences Institute, UMC Utrecht, Utrecht, The Netherlands Department of Radiology, UMC Utrecht, Utrecht, The Netherlands
Abstract. The amount of coronary artery calcification (CAC) is a strong and independent predictor of cardiovascular events. Non-contrast enhanced cardiac CT is considered a reference for quantification of CAC. Recently, it has been shown that CAC may be quantified in cardiac CT angiography (CCTA). We present a pattern recognition method that automatically identifies and quantifies CAC in CCTA. The study included CCTA scans of 50 patients equally distributed over five cardiovascular risk categories. CAC in CCTA was identified in two stages. In the first stage, potential CAC voxels were identified using a convolutional neural network (CNN). In the second stage, candidate CAC lesions were extracted based on the CNN output for analyzed voxels and thereafter described with a set of features and classified using a Random Forest. Ten-fold stratified cross-validation experiments were performed. CAC volume was quantified per patient and compared with manual reference annotations in the CCTA scan. Bland-Altman bias and limits of agreement between reference and automatic annotations were -15 (-198–168) after the first stage and -3 (-86 – 79) after the second stage. The results show that CAC can be automatically identified and quantified in CCTA using the proposed method. This might obviate the need for a dedicated non-contrast-enhanced CT scan for CAC scoring, which is regularly acquired prior to a CCTA scan, and thus reduce the CT radiation dose received by patients. Keywords: Automatic coronary artery calcium scoring, Cardiac CTA, Convolutional neural network, Random Forest.
1
Introduction
Cardiovascular disease (CVD) is the global leading cause of death. The amount of coronary artery calcification (CAC) is a strong and independent predictor of CVD events, which can be identified and quantified in cardiac CT [1]. In clinical practice, CAC is routinely quantified using non-contrast enhanced, calcium scoring CT (CSCT) [2]. Recently, it has been shown that CAC may also be quantified in contrast-enhanced cardiac CT angiography (CCTA). Consequently,
This work has been financially supported by PIE Medical Imaging.
c Springer International Publishing Switzerland 2015 N. Navab et al. (Eds.): MICCAI 2015, Part I, LNCS 9349, pp. 589–596, 2015. DOI: 10.1007/978-3-319-24553-9_72
590
J.M. Wolterink et al.
a dedicated CSCT scan, which is often routinely acquired prior to CCTA, might potentially be omitted. This could reduce the radiation dose of a typical cardiac CT examination by 40-50% [3]. CAC in CSCT can be identified manually by an expert or automatically [4,5]. In both situations, a threshold of 130 HU is used to identify connected voxels representing CAC. This method is not generalizable to CCTA. The coronary artery lumen is typically enhanced beyond 130 HU, which makes differentiation of CAC and lumen challenging. Other global attenuation thresholds for manual CAC scoring in CCTA have therefore been proposed [6,7].
Data Loading...
