Correction of Photon Attenuation and Collimator Response for a Body-Contouring SPECT/CT Imaging System

Correction of Photon Attenuation and Collimator Response for a Body-Contouring SPECT/CT Imaging System

Youngho Seo, PhD¹, Kenneth H. Wong, PhD², Mingshan Sun, MS³, Benjamin L. Franc, MD¹, Randall A. Hawkins, MD, PhD¹^,2 and Bruce H. Hasegawa, PhD¹^,2^,3

¹ Department of Radiology, University of California, San Francisco, California
² Joint Graduate Group in Bioengineering, University of California, San Francisco and Berkeley, California
³ Department of Nuclear Engineering, University of California, Berkeley, California

View larger version (16K):

[in a new window]

FIGURE 1. CT image of acrylic water-filled resolution phantom (left). Linear attenuation coefficients were extracted along a line from attenuation maps from GE Discovery VH (middle) and using effective energy method (Eq. 1; right).

View larger version (38K):

[in a new window]

FIGURE 2. Composite CT image representing sum of all 128 slices from anthropomorphic phantom (left) and contours representing actual and calculated camera trajectories (right).

View larger version (16K):

[in a new window]

FIGURE 3. SNRs are plotted against iteration number for SPECT images reconstructed with MLEM algorithm for data from uniform cylindric phantom filled with ^99mTc (left) and from anthropomorphic torso phantom filled with ¹¹¹In (right). Curves are shown for FBP, for MLEM with attenuation correction only (AC), and for MLEM with correction for collimator response (CDRC).

View larger version (45K):

[in a new window]

FIGURE 4. SPECT images (top) of uniform cylindric phantom and corresponding count profiles (bottom) along a line through center of image. SPECT images were reconstructed with FBP (left), with MLEM including collimator response correction but without attenuation correction (20 iterations, middle), and with MLEM including corrections for both collimator response and photon attenuation (40 iterations, right).

View larger version (20K):

[in a new window]

FIGURE 5. Target-to-background was analyzed using anthropomorphic phantom with ¹¹¹In-filled liver compartment containing 2 spheric lesions with higher concentrations. CT image (left) of phantom with corresponding SPECT images reconstructed using 40 iterations of MLEM with corrections for photon attenuation and collimator response from acquisitions with circular orbit (middle) and with body contouring (right). Circular ROI in hot spheric lesion was used to define target activity, whereas irregular ROI in liver compartment was used to define background activity.

View larger version (49K):

[in a new window]

FIGURE 6. SPECT/CT images from clinical ¹¹¹In-capromab pendetide study including CT (left) and SPECT images reconstructed using FBP (middle) and using 40 iterations of MLEM with correction for photon attenuation and collimator response (right). Images show ROIs to differentiate radioactivity concentration in bone marrow and in soft-tissue regions.

View larger version (107K):

[in a new window]

FIGURE 7. SPECT/CT images from clinical ¹¹¹In-capromab pendetide study show SPECT images reconstructed using MLEM without photon attenuation and collimator response corrections (10 iterations, top left), using MLEM with photon attenuation correction only (15 iterations, top middle), and using MLEM with collimator response correction only (20 iterations, top right), using MLEM with corrections for both photon attenuation and collimator response (40 iterations, bottom left), and with FBP (bottom right). Corresponding CT image is shown at bottom middle.