Semi-Automatic Graphical Tool for Measuring Coronary Artery Spatially Weighted Calcium Score from Gated Cardiac Computed Tomography Images

The Mission of the Biomedical Engineering and Imaging Institute is to create innovative biomedical engineering and imaging technologies that can be utilized in medical science and clinical practice. The goal is to identify and develop techniques early and accurate diagnosis of diseases, which can then be useful in optimizing more effective treatments. Our cardiovascular imaging group concentrates on image acquisition and analysis techniques for MRI, PET and CT imaging.

CT imaging can be used to measure calcification and determine the severity of coronary artery disease without invasive procedures. The Agatston score is the most commonly used method for calcium scoring, but it relies on signal thresholds, which can cause early stages of the disease to be missed due to small, lowly attenuating deposits not being detectable above image noise. Dr.Jason Liang and his colleagues at the University of Washington developed the idea of the spatially weighted calcium score.

However, at present, there are no reliable and consistent tools to compute the score. We have created a program that offers researchers the ability to calculate it accurately and repeatedly. The specially weighted calcium score is designed to be sensitive to detect micro calcification, which occurs before at advanced coronary calcification can be detected by traditional calcium scoring CT scans.

That sensitivity is significant in identifying subclinical coronary artery disease and by providing quantitative data, which can be useful when assessing novel risk factors, like metal exposure from e-cigarettes, evaluating modifiable risks in younger populations, and exploring the impact of interventions. The immediate application of this work is to examine the relationship between coronary artery calcification and exposure to toxic metals from use of e-cigarettes. This project is funded by NIH and LED by Professor Navas-Acien Ian at Columbia University.

To begin, create the appropriate folder structure for the protocol, navigate to the project's main folder and locate the program file. double-click on the file to launch the software and display the program's code. To launch the program, click anywhere on the editor window and click the green run button on the top ribbon of the editor tab.

Click the open DICOM button in the bottom left corner to access the file's directory. Navigate to the main project folder and find the original data. Select the patient's folder to be analyzed and click open.

The images in axial, sagittal and coronal views will appear on the screen. To examine the slices in a specific view, hover and scroll over them. Use the cross hairs to display the location of the pointer at that moment.

Using the sliders found in the bottom right corner of the program, adjust the brightness and contrast of the image. To analyze the coronary artery calcification, click the zero milligrams per milliliter phantom button. Hover over the axial view and scroll until the zero milligrams per milliliter section of the phantom is visible.

Position the cursor at the zero milligrams per milliliter phantom midpoint in the axial view. Observe the cross hairs in the sagittal and coronal views without moving the cursor. Scroll through several slices until the cross hairs and the three views are centered on the zero milligrams per milliliter phantom.

To generate a red cluster of circles in the axial view in a column of three points in the sagittal and coronal views, place a 10 by 10 point grid on the current and adjacent slices. If a phantom is unavailable or of poor quality, select the no phantom button to use an aggregation of 10 sample phantoms and their corresponding weighting functions. Then select an artery button to start tracing each of the four coronary arteries.

Hover over the axial view. Scroll to navigate to either the proximal or distal end of the chosen artery and observe the shape of the artery in the slice. To mark a circular artery, click on its center in the axial view and a five millimeter diameter point will appear, which will also be visible in the sagittal and coronal views.

If the artery is more visible in the sagittal or coronal views, place a point there ensuring it aligns with the artery's center in the axial view. Quickly, double-click in any of the three views to delete the suboptimal placed points. Observe the bottom left area for a notification confirming the points deletion.

If the shape of the artery is not circular in the axial plane, ensure the desired axial slice is in view and click the draw ROI button. In the popup window, scroll to zoom in or out and start tracing around the artery with single clicks. To delete the previous point in the artery tracing, use the backspace key while the ROI is open and close the ROI by double-clicking on the last or first point placed.

Refine the closed ROI by dragging its perimeter points or double-clicking on the perimeter to add a point. Then lock in and redraw, the ROI will appear at the bottom of the popup window. Once satisfied with the current ROI click the lock-in button and close the pop-up window using the red button in the top left corner.

To delete a locked-in ROI, double-click any perimeter points in the axial view of the main program window. Scroll through one slice in the axial view and repeat the procedure until the end of the chosen artery is reached. After completing tracing, re-click on the arteries button to ensure no accidental points were placed.

After tracing all four arteries, click the SWCS button in the bottom right corner to generate results. The bottom left area will show the progress and display, done processing, when finished. Close the program window by clicking the red button.

To access the results, open the files directory, navigate to the project's main folder, and enter the metadata folder. Look for a folder with the same name as the original folder for the subject, and find the file in the desired format. Review the PDFs to obtain the final SWCS and Agatston score for the case and the waiting function used.

The percentage difference between the Agatston score obtained from the program versus the commercial software is presented using the Bland-Altman plot.