Abstract: A method of analyzing computed tomography (CT) images comprises receiving an initial CT image of an object, identifying calcium-free regions and a calcified region in the initial CT image of the object, generating a calcium-free image patch, and applying the calcium-free image patch to the initial CT image patch to produce a final CT image. The initial CT image shows a calcium deposit and a target structure in the object. The calcified region in the initial CT image shows the calcium deposit in the object obscuring a portion of the target structure. The calcium-free regions show the remaining portions of the target structure. The calcium-free image patch corresponds to the calcified region in the initial CT image. The final CT image shows the calcium-free image patch and the calcium-free region from the initial CT image. The calcium-free image patch is generated and applied using a convolutional neural network.

Abstract: Calcific and noncalcific aortic tissue components can be quantified. Pre-intervention planning computed tomography angiography imaging data is received. A region of interest is defined between the lower coronary ostium and the virtual basal ring. Cross-sectional images of the region of interest are rendered and calcific and noncalcific tissue components are identified based on Hounsfield unit thresholds. The volumes of the identified calcific and noncalcific tissue components are calculated and used to determine a total tissue volume (e.g., fibrocalcific volume) for the valve, as well as component percentages of the total tissue volume for the calcific and noncalcific components. These volumes and/or component percentages can be leveraged to predict severe AS, identify prognosis of post-TAVI outcomes, or otherwise facilitate planning of medical intervention.

Abstract: Convolutional long short-term memory (LSTM) networks are leveraged to segment and/or quantify medical imaging data. A multi-branch architecture makes use of an attention branch and a main branch. The main branch includes a dense block followed by a segmentation head, and is configured to consider a single input slice of the imaging data. This main branch is able to segment larger and easier-to-classify targets. The attention branch, however, makes use of a sequential processor that includes a convolutional LSTM (ConvLSTM) followed by a segmentation head and an attention head. This attention branch is configured to process an input slice and additional adjacent slices. This attention branch is able to segment smaller and more difficult to classify targets.

Abstract: Calcific and noncalcific aortic tissue components can be quantified. Pre-intervention planning computed tomography angiography imaging data is received. A region of interest is defined between the lower coronary ostium and the virtual basal ring. Cross-sectional images of the region of interest are rendered and calcific and noncalcific tissue components are identified based on Hounsfield unit thresholds. The volumes of the identified calcific and noncalcific tissue components are calculated and used to determine a total tissue volume (e.g., fibrocalcific volume) for the valve, as well as component percentages of the total tissue volume for the calcific and noncalcific components. These volumes and/or component percentages can be leveraged to predict severe AS, identify prognosis of post-TAVI outcomes, or otherwise facilitate planning of medical intervention.

Abstract: A method of analyzing computed tomography (CT) images comprises receiving an initial CT image of an object, identifying calcium-free regions and a calcified region in the initial CT image of the object, generating a calcium-free image patch, and applying the calcium-free image patch to the initial CT image patch to produce a final CT image. The initial CT image shows a calcium deposit and a target structure in the object. The calcified region in the initial CT image shows the calcium deposit in the object obscuring a portion of the target structure. The calcium-free regions show the remaining portions of the target structure. The calcium-free image patch corresponds to the calcified region in the initial CT image. The final CT image shows the calcium-free image patch and the calcium-free region from the initial CT image. The calcium-free image patch is generated and applied using a convolutional neural network.

Abstract: A method of quantifying plaques imaged by cardiac computed tomography angiography (“CCTA”) scan data. Calcified and non-calcified component thresholds are determined based at least in part on attenuation values of a pool of blood in the CCTA scan data. An epicardial fat threshold is determined and used to classify epicardial fat in the CCTA scan data. A portion of CCTA scan data positioned between a detected outer boundary of the coronary artery and a portion classified as lumen is classified as arterial wall. NCP and CP seeds are identified in the arterial wall portion. Portions of the CCTA scan data continuous with a NCP seed and having attenuation values greater than an artery wall value and less than the NCP threshold are classified as NCP, and portions of the CCTA scan data continuous with the CP seed and having attenuation values greater than the CP threshold are classified as CP.

Abstract: A method of quantifying plaques imaged by cardiac computed tomography angiography (“CCTA”) scan data. Calcified and non-calcified component thresholds are determined based at least in part on attenuation values of a pool of blood in the CCTA scan data. An epicardial fat threshold is determined and used to classify epicardial fat in the CCTA scan data. A portion of CCTA scan data positioned between a detected outer boundary of the coronary artery and a portion classified as lumen is classified as arterial wall. NCP and CP seeds are identified in the arterial wall portion. Portions of the CCTA scan data continuous with a NCP seed and having attenuation values greater than an artery wall value and less than the NCP threshold are classified as NCP, and portions of the CCTA scan data continuous with the CP seed and having attenuation values greater than the CP threshold are classified as CP.