Umeå universitets logga

Positron Emission Tomography (PET) is an important tool for detection, staging and follow-up in a wide range of diseases, including cancer and neurological disorders. As a functional imaging tool, PET can visualize biological processes, where positron emitting radioactive isotopes are connected to molecules with different functions in the body. While PET-images can be visually interpreted, they can also be used for quantitative measurements, where functions such as glucose metabolism, dopamine receptor function, and blood-flow can be quantified. Measurements can be performed in static imaging, or in dynamic imaging where graphical methods can be used for analysis.

PET images benefit from fusion with anatomical images which facilitates the interpretation. The combination of PET with computed tomography (CT) as in PET/CT hybrid equipment is a well-established imaging method. Magnetic Resonance Imaging (MRI) has some advantages over CT such as the high soft tissue contrast, but the combination with PET in a fully integrated system is far more technically challenging. Most of the technical concerns have been solved, and PET/MRI modalities are now commercially available.

Among the remaining challenges, the attenuation correction is still not yet completely solved, where the attenuation maps on the PET/MRI modalities are approximate and bone is not accounted for in all parts of the body. There are also challenges with quantitative PET in general, where for example low spatial resolution and presence of noise can lead to quantitative errors. The purpose of this thesis was to investigate and develop strategies to reduce quantitative errors in PET imaging with special focus on PET/MRI.

In study I, we studied the limits for quantification of size and uptake in small lesions in PET images reconstructed with a resolution modelling algorithm. We constructed a phantom of small balloons and reconstructed images with three different algorithms and measured volume and activity concentration in the images. The measured activity concentration in the lesions was corrected for the low resolution that yields partial-volume effects (PVE). We found that resolution modelling improved quantification of all lesions, and that in combination with correction factors, lesions larger than ~9 mm diameter could be correctly quantified.

Study II is focused on the effect of frame time length on the graphical Logan-analysis for dynamic studies with 11C-raclopride. Logan analysis is reported to be sensitive to noise, and image noise is heavily dependent on the frame time length. Noise can also generate bias when using iterative reconstruction methods. Weivconcluded that with region-based analyses, a bias of approximately 10% in the non-displaceable binding potential was found when using the shortest time frames, and that the bias was mainly caused by the reconstruction algorithm. Long time frames generated stable parameters.

The last two studies focused on the attenuation correction in PET/MRI hybrid equipment. In study III, a method for attenuation correction in PET/MRI was implemented and evaluated. The method is developed for the pelvic region and is based on statistical decomposition of T2-weighted images. We found that the new method improved quantification, especially in regions in vicinity of bone. In study IV, we proposed a concept for patient-specific quality assurance of attenuation maps, based on measurements of the MRI B0-field. The method shows potential to find errors in the attenuation map related to metallic implants, air, and patient contour.

The work in this thesis has contributed to increased knowledge about the effect of resolution and noise for quantification in PET images. It has also introduced a new method for attenuation correction in PET/MRI, and a concept for quality assurance of PET/MRI attenuation maps.

Magnetic resonance imaging (MRI) offers high soft-tissue contrast compared to computed tomography (CT). This contrast is helpful in many cases, not least for delineating tumours for radiotherapy treatment, and has led to increasing use in radiotherapy treatment planning (RTP).

When RTP is based on CT images, the treatment planning system can get the approximate electron density of the tissues from the electron density equivalent information that constitutes the CT images. This information is needed to calculate the dose to the patient from radiotherapy. Therefore, for an MR-only workflow, the MRI image must be transformed into information that can yield electron density information. The predominant way is to convert the MRI images into CT-like images, also known as substitute CT (sCT).

Positron emission tomography (PET) imaging can benefit from being combined with anatomical imaging, and the PET/CT hybrid machine is well established. The soft tissue contrast properties of the MRI images are also valuable for a PET/MRI hybrid system. However, it also adds the option for simultaneous acquisition of PET and anatomical (MRI) images which is not feasible with CT images for a PET/CT system. However, the PET/MRI combination is more technically challenging. While most of the concerns have been solved or mitigated, and PET/MRI systems have been commercially available for some years, there are still outstanding issues. The attenuation maps used in the reconstruction of PET acquisitions are one of the concerns that have yet to be solved entirely. These attenuation maps in the PET/MRI systems are approximations where, e.g., bone is not fully accounted for in all parts of the body.

The MRI images suffer from geometric distortions dependent on the MRI scanner as well as the imaged patient itself. These distortions can affect the sCT conversion for RTP and the attenuation maps for PET reconstruction.

This thesis aimed to investigate the size of the geometric MRI distortions for different settings on the MRI scanner and their effect on the resulting RTP and reconstructed PET images. Such information can aid in optimising the MRI imaging for different purposes. It can also give some information needed to determine tests to run in a quality assurance (QA) regime.

In Paper I, we studied the machine-dependent MRI gradient-field nonlinearity distortions and their effect on PET reconstruction. We simulated different levels of incomplete corrections for gradient-field nonlinearities in CT images from PET/CT acquisitions. The resulting distorted images were then used for rerunning the reconstruction of PET data, and the effect on reconstructed standardised uptake value (SUV) was studied. We found that residual gradient-field nonlinearity dependent geometrical distortions of ±2.3 mm at 15 cm radius from the scanner isocenter lead to SUV quantification errors below 5%. This is also below the test-retest variability caused by instrumentation and intra-patient factors for PET/CT systems.

In Paper II, we developed a method for simulating the patient-induced susceptibility effect based on CT images. The method consisted of converting the CT images to magnetic susceptibility maps. These magnetic susceptibility maps were then used in the simulation by calculating the local shifts in the main magnetic field (B0). From these local shifts in B0, a displacement map was calculated, and this was, in turn, applied to the original CT images. The simulation was validated through comparisons between the simulation and analytical results for both a homogeneous sphere and a homogeneous cylinder. This method was tested on a set of eight prostate cancer patients. We found that setting the frequency encoding bandwidth to a minimum of twice the water-fat shift would keep the maximum distortion from the patient-induced susceptibility effect below 1 pixel. However, the required frequency encoding bandwidth was shown to be dependent on the imaged area, and lower bandwidth could, e.g., be used for the pelvic area.

The simulation method from Paper II was then used in Paper III, where we investigated the dosimetric impact of residual MRI system distortions, patient-induced susceptibility effects and patient-specific shimming. The latter was simulated using an in-house Matlab algorithm. The residual system distortions were determined using phantom measurements. These distortions were then combined with a simulated patient-induced susceptibility effect and patient-specific shimming. The combined distortions and patient-specific shimming were then applied to patient CT images. The distorted patient images were then used for RTP, and the resulting treatment plan was transferred to the original patient CT datasets and recalculated. This allowed for isolating the effect of the applied distortions on the dose distribution. We concluded that the dosimetric impact of MRI distortions within the target volume and nearby organs at risk is small for high bandwidth spin echo sequences. We also saw a worsening in field variations for the user-defined region ofinterest shimming.

Paper IV presented a proof of concept for patient-specific QA of sCTs and attenuation maps. In this study, we compared measured B0-maps with ones simulated using Paper II’s method. The simulations were based on sCT images from MRI acquisitions from the same imaging session as the measured B0-maps. This method shows potential for identifying errors or problematic areas of sCTs and attenuation maps. It should also be feasible to make the method fast enough to use while the patient is still in the scanner so that images could be retaken without having to recall thepatient if a problem is detected.

This work has contributed to the knowledge and methods needed for the necessary considerations for optimisation and setting up a QA protocol aimed at PET/MRI and MR-only radiotherapy.