Katarina Sjögreen

3D absorbed dose calculations based on SPECT: Evaluation for 111-In/90-Y therapy using Monte Carlo simulations

A general method is presented for patient-specific three-dimensional (3D) absorbed dose calculations based on quantitative SPECT activity measurements. The computational scheme includes a method for registration of the CT study to the SPECT image, and compensation for attenuation, scatter, and collimator-detector response including septal penetration, performed as part of an iterative reconstruction method. From SPECT images, the absorbed dose rate is calculated using an EGS4 Monte Carlo code, which converts the activity distribution to an absorbed dose rate distribution. Evaluation of the accuracy in the activity quantification and the absorbed dose calculation is based on realistic Monte Carlo simulated SPECT data of a voxel-computer phantom and (111)In and (90)Y. Septal penetration was not included in this study. The SPECT-based activity concentrations and absorbed dose distributions are compared to the actual values; the results imply that the corrections for attenuation and scatter yield results of high accuracy. The presented method includes compensation for most parameters deteriorating the quantitative image information. Inaccuracies are, however, introduced by the limited spatial resolution of the SPECT system, which are not fully compensated by the collimator-response correction. The proposed evaluation methodology may be used as a basis for future inter-comparison of different dosimetry calculation schemes.

High resolution pinhole SPECT for tumor imaging

High-resolution, non-invasive, 3D-imaging techniques would greatly benefit the investigation of the localization properties of tumor-specific radiopharmaceuticals in laboratory animals. The present study reports how pinhole SPECT can be applied to tumor localization studies in small laboratory animals to provide high resolution SPECT images in vivo. Pinhole SPECT was performed using a rotating scintillation camera, equipped with a pinhole collimator. The sensitivity of a 2 mm diameter collimator at 45 mm from the source is 90 cps/MBq for 99mTc. The planar spatial resolution at a 45 mm distance is 2.2 mm. The transaxial spatial resolution, with a distance of 45 mm between the collimator aperture and the axis of rotation, is 3.1 mm. For SPECT imaging, spatial linearity is preserved across the usable field-of-view. The major advantage of the high resolution properties of pinhole tomography is demonstrated by the enhanced lesion-to-normal-brain uptake ratio achieved on tomographic slices as compared to planar images. For example, 201Tl tumor-to-normal-brain uptake ratios of 1.1 to 1.3 observed on planar images, corresponded to ratios ranging from 3.2 to 3.7 on the SPECT slices. Examples of the activity distributions of two radiopharmaceuticals in tumor and in normal brain for sagittal and coronal images are given. In all cases, tumors are clearly delineated on the pinhole SPECT slices. The present study shows that pinhole SPECT performed with standard SPECT instrumentation can give high spatial resolution images, with a FWHM approximately 3 mm and a sensitivity approximately 100 cps/MBq for 99mTc.

Boundary finding using Fourier surfaces of increasing order [simulated medical images]

Boundary finding in simulated medical images is performed by optimizing the Fourier coefficients in a parametric surface representation with respect to an objective function. The deformable model is fitted to the data using the brightness gradient component which is normal to the surface. A low order (<10) Fourier series expansion offers a sufficiently accurate representation for many inherently smooth objects that occur in medical imaging. Experimental results are presented for simulated image objects corresponding to organs of the anthropomorphic Zubal phantom. Two different optimization methods are studied concerning robustness and computational efficiency. The effect of increasing the Fourier expansion order is investigated for various noise levels.

Registration of Abdominal CT and SPECT Images Using Compton Scatter Data

The present study investigates the possibility to utilize Compton scatter data for registration of abdominal SPECT images. A method for registration to CT is presented, based on principal component analysis and cross-correlation of binary images representing the interior of the patient. Segmentation of scatter images is performed with two methods, thresholding and a deformable contour method. To achieve similarity of organ positions between scans, a positioning device is applied to the patient. Evaluation of the registration accuracy is performed with a) a 131I phantom study, b) a Monte Carlo simulation study of an anthropomorphic phantom, and c) a 123I patient trial. For a) r.m.s. distances between positions that should be equal in CT and SPECT are obtained to 1.0±0.7 mm, which thus for a rigid object is at sub pixel level. From b) results show that r.m.s. distances depend on the slice activity distribution. With a symmetrical distribution deviations are in the order of 5 mm. In c) distances between markers on the patient boundary are at the maximum 16 mm and on an average 10 mm. It is concluded that by utilizing the available Compton scatter data, valuable positioning information is achieved, that can be used for registration of SPECT images.

Deformable Fourier surfaces for volume segmentation in SPECT

Three-dimensional boundary finding based on Fourier surface optimization is presented as a method for segmentation of SPECT images. Being robust against noise and adjustable with respect to its detail resolution, it forms an interesting alternative in this application area. A three-dimensional approach can also be assumed to increase the possibility of delineating low contrast regions, as compared to a two-dimensional slice-by-slice approach. We apply boundary finding to Monte Carlo simulated SPECT images of the computer-based anthropomorphic Zubal phantom in order to evaluate the influence of object contrast and noise on the segmentation accuracy. Segmentation is also performed in real patient images.

Registration of whole-body scintillation camera images for conjugate view quantification

A registration method for whole-body (WB) scintillation camera images is presented, with application to conjugate view activity quantification. Accurate image registration is here required if the quantification is performed on an image basis, for (a) serial emission images, for analysis of the biokinetics, and (b) transmission and emission images, for a pixel-based attenuation correction. The spatial transformation used for registration is composed of both global and local transformations, including rigid, projective and one curved transformation. As similarity measure, the mutual information is used. An initial coarse solution is first calculated by cross-correlation. Optimization is then performed in a sequence, beginning with the legs position, followed by the abdomen and head. Evaluation has been performed for Monte Carlo simulated 131-I images of an anthropomorphic WB phantom. Obtained registration errors are within one pixel (<3.6 mm).