Bioresorbable Scaffold Visualization in IVOCT Images Using CNNs and Weakly Supervised Localization

摘要

Bioresorbable scaffolds have become a popular choice for treatment of coronary heart disease, replacing tradi-tional metal stents. Often, intravascular optical coherence tomography is used to assess potential malappositionafter implantation and for follow-up examinations later on. Typically, the scaffold is manually reviewed by anexpert, analyzing each of the hundreds of image slices. As this is time consuming, automatic stent detection andvisualization approaches have been proposed, mostly for metal stent detection based on classic image processing.As bioresorbable scaffolds are harder to detect, recent approaches have used feature extraction and machinelearning methods for automatic detection. However, these methods require detailed, pixel-level labels in eachimage slice and extensive feature engineering for the particular stent type which might limit the approaches'generalization capabilities. Therefore, we propose a deep learning-based method for bioresorbable scaffold visu-alization using only image-level labels. A convolutional neural network is trained to predict whether an imageslice contains a metal stent, a bioresorbable scaffold, or no device. Then, we derive local stent strut informa-tion by employing weakly supervised localization using saliency maps with guided backpropagation. As saliencymaps are generally diffuse and noisy, we propose a novel patch-based method with image shifting which allowsfor high resolution stent visualization. Our convolutional neural network model achieves a classification accuracyof 99:0% for image-level stent classification which can be used for both high quality in-slice stent visualizationand 3D rendering of the stent structure.