Tbased Attenuation Correction Transforming Attenuation Coefficients

In addition to acquiring co-registered anatomical and functional images, a further advantage of the combined PET/CT scanner is the potential to use the CT images for attenuation correction of the PET emission data, eliminating the need for a separate, lengthy PET transmission scan. The use of the CT scan for attenuation correction not only reduces whole-body scan times by at least 30% [27], but provides essentially noiseless attenuation correction factors compared to those from a standard PET transmission scan. CT-based attenuation values are, however, energy dependent, and hence the correction factors derived from a CT scan at a mean photon energy of 70 keV must be scaled to the PET energy of 511 keV, for which a hybrid scaling algorithm was developed [28]. This scaling approach is based on previous work by La Croix and Tang [29, 30] who have shown that attenuation correction factors for SPECT emission data can be derived from complementary CT transmission data. The single scale-factor approach works well for soft tissues, but serious overestimation of the attenuation properties of cortical bone and ribs is observed, especially with increasing difference of the transmission and emission

Figure 8.4. Male patient with metastatic melanoma before (a) and after (b) chemotherapy in 11/98 and 12/98, respectively. PET/CT images are from the prototype PET/CT. Multiple lesions are depicted as FDG avid and localized accurately within the anatomy of the patient. Each whole-body scan took about 1 h. PET/CT images before and after therapy are registered by hand and selected axial views are shown to demonstrate multi-variant response to therapy.

Figure 8.4. Male patient with metastatic melanoma before (a) and after (b) chemotherapy in 11/98 and 12/98, respectively. PET/CT images are from the prototype PET/CT. Multiple lesions are depicted as FDG avid and localized accurately within the anatomy of the patient. Each whole-body scan took about 1 h. PET/CT images before and after therapy are registered by hand and selected axial views are shown to demonstrate multi-variant response to therapy.

photon energies [29]. Nevertheless, the overestimation of attenuation in bone translated into only a minor average overestimation of the tracer uptake in the corrected SPECT images due to the low fraction of voxels containing bone compared to other tissues.

In anticipation of the prototype PET/CT the original scaling approach was extended to CT-based attenuation correction of PET emission data [28]. A bi-linear scaling is employed to account for both the photon energy difference between CT and PET, and the different attenuation properties of low-Z (soft tissues) and high-Z (bone) materials in the range of the lower energy X-ray photons (Fig. 8.5).

The algorithm is based on the observation that for water, lung, fat, muscle and other soft tissues, the mass attenuation coefficient (linear attenuation coefficient divided by density) at CT and PET energies is approximately the same. While the actual value is different at the effective CT energy of 70 keV to that at 511 keV, all the tissues can be scaled with a single factor: the ratio of the mass attenuation coefficient at 511 keV to that at 70 keV. The exception is bone because, at CT energies, the mass attenuation coefficient is somewhat higher than that for the other tissues due to the increased photoelectric contribution from calcium. The ratio of the mass attenuation at 511 keV to that at 70 keV is approximately 0.53 for soft tissues, and 0.44 for bone.

CT-based PET attenuation correction factors are generated in a 4-step procedure [28]:

1. the CT images are divided into regions of pixels classified as either non-bone or bone by simple thresholding. A threshold at 300 Hounsfield Units (HU) separates spongiosa and cortical bone from other tissues,

2. the pixel values in the CT image in HU are converted to attenuation coefficients of tissue (^T) at the effective CT energy (~70 keV) using the expression: = ^W(HU/1000+1); is the attenuation coefficient for water,

3. the non-bone classified pixel values are then scaled with a single factor of 0.53, and bone classified pixel values are scaled with the smaller scaling factor of 0.44,

4. attenuation correction factors are generated by integrating (forward projecting) along coincidence lines-of-response through the segmented and scaled CT images, with the CT spatial resolution degraded to match that of the PET. Oblique lines-of-response are obtained in the same way by integration through the CT volume.

Today bi-linear scaling methods [28, 31] are widely accepted for clinical PET/CT imaging, and are, with minor modifications, used routinely for CT-based attenuation correction of the PET emission data [32, 33].

Mass Attenuation Coefficient Tissue

Figure 8.5. Mass attenuation coefficients for soft tissue and bone (a) differ significantly for lower photon energies for which the photoelectric effect is the dominant interaction with matter. The bi-linear scaling and segmentation approach (b) accounts for the different attenuation properties of soft tissue and bone by, first, segmenting (Se) the CT, and, second, by applying a tissue dependent scale factor (Sc) to these pixels. The greatly increased photon flux used in CT results in essentially noiseless attenuation maps (c) and in attenuation correction factors compared with those derived from a standard PET transmission scans (d).

Figure 8.5. Mass attenuation coefficients for soft tissue and bone (a) differ significantly for lower photon energies for which the photoelectric effect is the dominant interaction with matter. The bi-linear scaling and segmentation approach (b) accounts for the different attenuation properties of soft tissue and bone by, first, segmenting (Se) the CT, and, second, by applying a tissue dependent scale factor (Sc) to these pixels. The greatly increased photon flux used in CT results in essentially noiseless attenuation maps (c) and in attenuation correction factors compared with those derived from a standard PET transmission scans (d).

The time for the acquisition of the attenuation data for a whole-body study can be reduced to one minute, or less, by using a fast CT scan instead of a lengthy PET transmission measurement. Furthermore CT transmission data acquired in post-injection scenarios are not noticeably affected by the emission activity inside the patient due to the high X-ray photon flux [22]. Therefore corrective data processing as in post-injection PET transmission imaging [34,35] is not required.

While, in principle, the CT-based attenuation coefficients are unbiased and essentially noiseless, there are a number of practical limitations. These include respiration effects, truncation of the CT field-of-view when imaging with the arms down, and the effect of using CT contrast.

How To Prevent Skin Cancer

How To Prevent Skin Cancer

Complete Guide to Preventing Skin Cancer. We all know enough to fear the name, just as we do the words tumor and malignant. But apart from that, most of us know very little at all about cancer, especially skin cancer in itself. If I were to ask you to tell me about skin cancer right now, what would you say? Apart from the fact that its a cancer on the skin, that is.

Get My Free Ebook


Post a comment