NHLBI-Emory Advanced Cardiac CT Reconstruction

Overview

Background: Doctors use computed tomography (CT) to get detailed pictures of the heart. CT uses x-rays to gather raw data. Computers assemble this data to make the images doctors look at. A new computer technique can make higher resolution images from the same CT scans. In this natural history study, researchers will take normal CT images of the heart. They will compare those images to super high-resolution (super high-res) images made with a super-computer. Objective: To improve the quality of heart CT scans by using new methods to create the images. Eligibility: People aged 18 years or older who need a CT scan for heart disease. Design: Participants will have a normal CT scan. A substance will be injected through a tube in their arm. They will lie on a table in a large, donut-shaped machine. An X-ray tube will move around their body, taking many pictures. Researchers will use the normal CT scans to create super high-res images. They may do this at the NIH. They may also send the images to the company that made the CT scanner. Participants personal information will be removed before images are sent to the company. The personal information will be replaced by a code. The super high-res images will be returned to the NIH. Some information will be collected from participants medical records. Researchers will compare the normal scans to the super high-res images. Participants' own doctors will also have a chance to see the super high-res images. Participants' CT pictures will be stored and used for future NIH research.

Study Description: This is an observational technical development study of advanced mathematical image reconstruction of raw CT data on patients with structural heart disease, in order to create better pictures. The raw CT data are obtained in the course of clinically-indicated contrast-enhanced time-resolved cardiac CT without any change in clinical practice. The raw data are acquired on a specific imaging system (Canon Aquilion One) at the enrolling site(s). Subjects are recruited among candidates selected clinically to undergo time-resolved cardiac CT in the diagnosis and treatment of structural heart disease. After consenting in writing, data are abstracted from their medical record, and the raw CT data used for standard medical care are copied. There is no investigational component to the standard clinical CT data acquisition and there is no change in radiation exposure from standard clinical CT examination. Subject participation concludes at the end of the standard clinical CT examination. The raw CT data are sent with PII to NHLBI investigators who will remove identifiers and replace with a code before transmitting the image to the industry collaborator (Canon) for investigational advanced mathematical image reconstruction. The reconstructed CT images are then transmitted back to the NHLBI where the images are re-identified. The NHLBI investigators will then forward the reconstructed image to the site investigators. The advanced image reconstruction may be performed at NHLBI once adequate computer resources are available. The research analysis consists of a comparison of standard CT images from the enrolling site with the advanced mathematical image reconstructions ("investigational images") for quality metrics. The investigational advanced mathematical image reconstruction kernel may be changed iteratively for the purpose of technical development during this study. The advanced image reconstruction is considered investigational because it is a commercial device not yet cleared for marketing by the US Food and Drug Administration. The advanced image reconstruction yields images with four-fold higher spatial resolution that standard images, based on the same standard CT acquisition. The advanced reconstruction images are labeled investigational and offer potential benefit to research subjects who are also patients. Therefore, the images are labeled investigational and returned to the local physician-investigators, to use at their discretion. This includes the possibility of informing medical decision-making based on investigational higher-resolution images. Objectives: The objective of this research protocol is to determine whether advanced mathematical image reconstructions create superior images compared with standard cardiac CT image reconstruction in patients with structural heart disease. Endpoints: The primary endpoint is a semi-quantitative measure of superiority of advanced mathematical image reconstruction compared with standard cardiac CT image reconstruction for study-specific fine detail. Secondary endpoints include quantitative signal-to-noise (SNR), contrast-to-noise (CNR), and image noise. Exploratory endpoints include Likert-type scales of image noise, artifacts, subjective image contrast, image sharpness, and diagnostic confidence.