Early detection and diagnosis of coronary artery disease (CAD) could save lives and reduce healthcare costs. In this study, we propose a 3D Resnet-50 deep learning model to directly classify normal subjects and CAD patients on computed tomography coronary angiography images. Our proposed method outperforms a 2D Resnet-50 model by 23.65%. Explainability is also provided by using a Grad-GAM. Furthermore, we link the 3D CAD classification to a 2D two-class semantic segmentation for improved explainability and accurate abnormality localisation.