Correction of nonuniform attenuation and image fusion in SPECT imaging by means of separate X-ray CT

Abstract

Improvements in image quality and quantitation measurement, and the addition of detailed anatomical structures are important topics for single-photon emission tomography (SPECT). The goal of this study was to develop a practical system enabling both nonuniform attenuation correction and image fusion of SPECT images by means of high-performance X-ray computed tomography (CT). A SPECT system and a helical X-ray CT system were placed next to each other and linked with Ethernet. To avoid positional differences between the SPECT and X-ray CT studies, identical flat patient tables were used for both scans; body distortion was minimized with laser beams from the upper and lateral directions to detect the position of the skin surface. For the raw projection data of SPECT, a scatter correction was performed with the triple energy window method. Image fusion of the X-ray CT and SPECT images was performed automatically by auto-registration of fiducial markers attached to the skin surface. After registration of the X-ray CT and SPECT images, an X-ray CT-derived attenuation map was created with the calibration curve for 99mTc. The SPECT images were then reconstructed with scatter and attenuation correction by means of a maximum likelihood expectation maximization algorithm. This system was evaluated in torso and cylindlical phantoms and in 4 patients referred for myocardial SPECT imaging with Tc-99m tetrofosmin. In the torso phantom study, the SPECT and X-ray CT images overlapped exactly on the computer display. After scatter and attenuation correction, the artifactual activity reduction in the inferior wall of the myocardium improved. Conversely, the incresed activity around the torso surface and the lungs was reduced. In the abdomen, the liver activity, which was originally uniform, had recovered after scatter and attenuation correction processing. The clinical study also showed good overlapping of cardiac and skin surface outlines on the fused SPECT and X-ray CT images. The effectiveness of the scatter and attenuation correction process was similar to that observed in the phantom study. Because the total time required for computer processing was less than 10 minutes, this method of attenuation correction and image fusion for SPECT images is expected to become popular in clinical practice.

Clinical significance of CT density-based, non-uniform photon attenuation correction of deep-inspiratory breath-hold perfusion SPECT

“…Using the software on the image data processor, AC of original 128 9 128 matrix DIBrH perfusion SPECT at each lung section was automatically performed based upon 168 9 168 matrix co-registered CT-derived attenuation coefficient maps with the calibration curve for 99m Tc according to the previous studies [3,5,7,8,11]. These attenuation coefficient maps were created from the data of variable-effective linear-attenuation coefficient values calculated from spatially corrected CT data: attenuation coefficient = 0.…”

“…Since photon interaction is significant and complicated in the thorax with variable photon attenuation structures such as chest wall, vertebrae and lungs [1][2][3][4][5][6][7][8][9][10][11][12], non-uniform photon attenuation correction (AC) is indispensable for better assessment of lung perfusion on perfusion SPECT. The use of thoracic transmission computed tomography (TCT) obtained by radioactive external sources is one of the options for non-uniform AC of perfusion SPECT, but it has not been widely used for clinical practice, because of complicated procedure and unavailability of radioactive external sources in some SPECT systems [1,2,4,9,10].…”

Impaired Value of 99m Tc-GSA Scintigraphy as an Independent Risk Factor for Posthepatectomy Liver Failure in Patients with Hepatocellular Carcinoma

“…A SPECT/PET system using 99m Tc-GSA (Forte; Philips/ADAC Laboratories, Milpitas, CA, USA) was used in the study, as reported previously [20, 27]. The details of our protocol were written in the previous paper [20].…”

Imaging-Based Liver Function Tests – Past, Present and Future

“…High spatial and temporal resolution is necessary for the use of a multi-compartment model since the input function can only be determined from the portal vein and liver artery in this way. A further disadvantage of nuclear medicine techniques is the necessary and complicated attenuation correction and the resulting signal inhomogeneity [82]. An advantage of nuclear medicine techniques is the lack of a background signal.…”