Effects of movement artifacts in nuclear hybrid modalities for image diagnostic

Abstract

In the obtention of medical images, the patients' movement can modify the identification of the body components in an image. The combination of imaging techniques may not always be a solution to improve the imaging quality; therefore, an artifact analysis is commonly required prior to applying an imaging procedure in patients. In this work, we systematically evaluated the movements' artifacts caused by the patients' breathing during the images acquisition and their impact on the fusion of SPECT and CT modalities. We used a specific phantom placed on a platform to emulate the respiratory movement, finding artifacts not appreciable under the standard condition used to obtain the SPECT images due to its low spatial resolution. The artifacts produced a deformation of elements on the images. Therefore, image processing was necessary to identify the registration accuracy with SPECT and CT modalities in two states (phantom at rest and for a phantom with simulated respiratory movements). A systematic difference was obtained for the first case (11.7 mm), and a range of (7.4 mm to 16.1 mm) for the second one. For the volumes' evaluation, the optimal threshold value for CT was 0.40 and for SPECT was 0.25, giving a rapid solution to reduce the artifacts' impact on medical images.