Premium
WE‐D‐I‐609‐03: Improving the Quantitative Accuracy of a Dedicated Small Animal SPECT/CT Scanner
Author(s) -
Hwang A,
Taylor C,
Seo Y,
Iwata K,
Li J,
Patt B,
Hasegawa B
Publication year - 2005
Publication title -
medical physics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.473
H-Index - 180
eISSN - 2473-4209
pISSN - 0094-2405
DOI - 10.1118/1.1998548
Subject(s) - imaging phantom , correction for attenuation , calibration , attenuation , scanner , nuclear medicine , image resolution , single photon emission computed tomography , iterative reconstruction , scintillation , pixel , gamma camera , physics , optics , detector , computer science , artificial intelligence , medicine , quantum mechanics
Purpose: High‐resolution radionuclide imaging is an important method for noninvasive assessments of small animal models of human disease. To improve the quantitative accuracy of small animal SPECT/CT, we have developed methods to calibrate the imaging geometry and energy response, and to perform attenuation correction. Method and Materials: All studies were performed using the X‐SPECT (Gamma Medica, Inc.), which consists of a compact scintillation camera and microCT system on a common gantry. Attenuation correction was performed developing a method to convert CT image values into attenuation coefficients by imaging a calibration phantom containing materials having known linear attenuation coefficients, and was assessed by acquiring SPECT images of phantom containing an aqueous solution of iodine‐125. The pinhole imaging geometry parameters were determined by scanning three point sources, and using a fitting algorithm to determine the values of the parameters. A pixel by pixel energy calibration was performed by acquiring data in list mode from flood phantoms. Results: CT calibration curves were obtained showing the correlation between CT image intensity and the linear attenuation coefficient for photons emitted by iodine‐125 and technetium‐99m. SPECT data reconstructed with attenuation correction improved uniformity, by eliminating the cupping artifact that otherwise decreases image intensity at the image center by 30%. The calibration for the imaging geometry resulted in a 10% change in image dimensions compared to images reconstructed using the nominal values for the geometric parameters. The energy calibration corrected for photopeak changes that varied as a function of spatial position and radionuclide photon energy, and produced images with improved uniformity. Conclusion: We are able to improve the quality and quantitative accuracy of SPECT images by applying improved image reconstruction and list mode processing techniques. Conflict of Interest: Bruce Hasegawa receives research support from Gamma Medica. Brad Patt, Joshua Li, and Koji Iwata are Gamma Medica employees.