Electrocardiogram-correlated image reconstruction from subsecond spiral computed tomography scans of the heart

Abstract

Subsecond computed tomography (CT) scanning offers potential for improved heart imaging. We therefore developed and validated dedicated reconstruction algorithms for imaging the heart with subsecond spiral CT utilizing electrocardiogram (ECG) information. We modified spiral CT z-interpolation algorithms on a subsecond spiral CT scanner. Two new classes of algorithms were investigated: (a) 180 degrees CI (cardio interpolation), a piecewise linear interpolation between adjacent spiral data segments belonging to the same heart phase where segments are selected by correlation with the simultaneously recorded ECG signal and (b) 180 degrees CD (cardio delta), a partial scan reconstruction of 180 degrees + delta with delta < fan angle, resulting in reduced effective scan times of less than 0.5 s. Computer simulations as well as processing of clinical data collected with 0.75 s scan time were carried out to evaluate these new approaches. Both 180 degrees CI and 180 degrees CD provided significant improvements in image quality. Motion artifacts in the reconstructed images were largely reduced as compared to standard spiral reconstructions; in particular, coronary calcifications were delineated more sharply and multiplanar reformations showed improved contiguity. However, new artifacts in the image plane are introduced, mostly due to the combination of different data segments. ECG-oriented image reconstructions improve the quality of heart imaging with spiral CT significantly. Image quality and the display of coronary calcification appear adequate to assess coronary calcium measurements with conventional subsecond spiral CT.