Investigation of MR image distortion for radiotherapy treatment planning of prostate cancer

MR imaging based treatment planning for radiotherapy of prostate cancer is limited due to MR imaging system related geometrical distortions, especially for patients with large body sizes. On our 0.23 T open scanner equipped with the gradient distortion correction (GDC) software, the residual image distortions after the GDC were <5 mm within the central 36 cm x 36 cm area for a standard 48 cm field of view (FOV). In order to use MR imaging alone for treatment planning the effect of residual MR distortions on external patient contour determination, especially for the peripheral regions outside the 36 cm x 36 cm area, must be investigated and corrected. In this work, we performed phantom measurements to quantify MR system related residual geometric distortions after the GDC and the effective FOV. Our results show that for patients with larger lateral dimensions (>36 cm), the differences in patient external contours between distortion-free CT images and GDC-corrected MR images were 1-2 cm because of the combination of greater gradient distortion and loss of field homogeneity away from the isocentre and the uncertainties in patient setup during CT and MRI scans. The measured distortion maps were used to perform point-by-point corrections for patients with large dimensions inside the effective FOV. Using the point-by-point method, the geometrical distortion after the GDC were reduced to <3 mm for external contour determination and the effective FOV was expanded from 36 cm to 42 cm.     

Radiotherapy planning of the pelvis using distortion corrected MR images: the removal of system distortions

Image distortion is an important consideration in the use of magnetic resonance (MR) images for radiotherapy planning. The distortion is a consequence of system distortion (arising from main magnetic field inhomogeneity and nonlinearities in the applied magnetic field gradients) and of effects arising from the object/patient being imaged. A two stage protocol has been developed to correct both system and object-induced distortion in pelvic images which incorporates measures to maintain the quality, accuracy and consistency of the imaging and correction procedures. The first stage of the correction procedure is described here and involves the removal of system distortion. Object- (patient-) induced effects will be described in a subsequent work. Images are acquired with the patient lying on a flat rigid bed, which reproduces treatment conditions. A frame of marker tubes surrounding the patient and attached to the bed provides quality assurance data in each image. System distortions in the three orthogonal planes are mapped using a separate phantom, which fits closely within the quality control frame. Software has been written which automates the measurement and checking of the many marker positions which the test objects generate and which ensures that patient data are acquired using a consistent imaging protocol. Results are presented which show that the scanner and the phantoms used in measuring distortion give highly reproducible results with mean changes of the order of 0.1 mm between repeated measurements of marker positions in the same imaging session. Effective correction for in plane components of system distortion is demonstrated.     

Real-time correction of magnetic field inhomogeneity-induced image distortions for MRI-guided conventional and proton radiotherapy

Image-guided radiotherapy has the potential to increase the success of treatment by decreasing uncertainties concerning tumour position and shape. We pursue integrated diagnostic quality MRI functionality with radiotherapy systems to boost the possibilities of image guidance by providing images with superior soft-tissue contrast during treatment. However, the use of MR images in radiotherapy can be hindered by geometrical distortions due to magnetic field inhomogeneity problems. A method for fast correction of these distortions is presented and implemented. Using a 20 cm square phantom containing a regular grid, a measure of residual deformation after correction is established. At very low gradient strength (which leads to large deformations) a maximum displacement of 2.9 mm is shown to be reduced to 0.63 mm. Next, the method is applied in vivo to the case of pelvic body contour extraction for prostate radiotherapy treatment planning. Here, again with low gradient strengths, distortions of up to 6 mm can be reduced to 2 mm. All results are provided within a lag time of 8 ms. We discuss implications of image distortions for MRI-guided photon and proton radiotherapy separately, since the dose-depth curves in these treatments are very different. We argue that, although field inhomogeneities cannot be prevented from occurring, distortion correction is not always necessary in practice. This work opens new possibilities for investigating on-line MRI-based plan adaptations and ultimately MRI-based treatment planning.     

A study of artefacts in simultaneous PET and MR imaging using a prototype MR compatible PET scanner.

We have assessed the possibility of artefacts that can arise in attempting to perform simultaneous positron emission tomography (PET) and magnetic resonance imaging (MRI) using a small prototype MR compatible PET scanner (McPET). In these experiments, we examine MR images for any major artefacts or loss in image quality due to inhomogeneities in the magnetic field, radiofrequency interference or susceptibility effects caused by operation of the PET system inside the MR scanner. In addition, possible artefacts in the PET images caused by the static and time-varying magnetic fields or radiofrequency interference from the MR system were investigated. Biological tissue and a T2-weighted spin echo sequence were used to examine susceptibility artefacts due to components of the McPET scanner (scintillator, optical fibres) situated in the MR field of view. A range of commonly used MR pulse sequences was studied while acquiring PET data to look for possible artefacts in either the PET or MR images. Other than a small loss in signal-to-noise using gradient echo sequences, there was no significant interaction between the two imaging systems. Simultaneous PET and MR imaging of simple phantoms was also carried out in different MR systems with field strengths ranging from 0.2 to 4.7 T. The results of these studies demonstrate that it is possible to acquire PET and MR images simultaneously, without any significant artefacts or loss in image quality, using our prototype MR compatible PET scanner.     

Ultrasonic motor-induced geometric distortions in magnetic resonance images

Ultrasonic motors (USMs) are common actuators that can be safely used in the magnetic resonance imaging (MRI) environment. However, lack of MRI compatibility results in issues such as image distortion. This fact led researchers to shift focus from USMs to pneumatic and hydraulic actuators in development of surgical robots. The aim is to quantify and compensate the geometric distortion of MR images as generated by the presence of USMs. An ultrasonic motor was positioned in three orientations with respect to the bore axis. The induced distortions were compared across four image sequences. To reduce the distortions, three artifact reduction methods were employed. Geometric distortion is the only artifact in image slices farther from the motor. The various motor orientations lead to different distortions, with the lowest distortion for the z orientation. The maximum measured distortion of ten pixels occurred. This maximal distortion is equal to a 1-cm displacement of the displayed points relative to their actual locations and it is beyond the acceptable level for medical display standards. Bandwidth reduction reduced the distortion, with a 50% reduction for a doubled bandwidth. In conclusion, USMs can be preferred alternative because accurate targeting of pathologies can occur in free distorted images.     

Characterization of the susceptibility artifact around a prostate brachytherapy seed in MRI

Magnetic distortions surrounding a typical brachytherapy seed (IMC6711, OncoSeed) within a clinical magnetic resonance imager were modeled for a number of different seed orientations with respect to the main magnetic field. From these distortion maps, simulated images were produced. The simulated images were then compared to images experimentally acquired using a spin echo technique on a Philips 1.5 T magnetic resonance imaging scanner. The modeled images were found to conform very well to those acquired experimentally, thus allowing one to establish where the seed is positioned within the complex image distortion patterns. The artifact patterns were dependent on the orientation of the seed with the main magnetic field, as well as the direction of the read encode gradient. While all imaging schemes which employ a unidirectional linear read encode trajectory should produce the artifacts modeled in this article, sequences other than spin echo may produce additional artifacts. Gradient echo and steady-state free precession imaging techniques were also performed on the seed for comparison.     

Image distortion in MRI-based polymer gel dosimetry of gamma knife stereotactic radiosurgery systems

We have studied effects of MR (magnetic resonance) image distortion on polymer gel dosimetry of Gamma Knife stereotactic radiosurgery systems. MR images of BANG polymer gel phantoms were acquired by using a Hahn spin-echo sequence and a fast 3D imaging GRASS sequence. Image artifacts were studied by varying the directions of frequency encoding and the receiver bandwidth. The phantoms were also CT (computed tomography) scanned. The studies showed that the measured dose distributions are shifted by 1.8+/-0.5 mm (2 pixels) in the frequency encoding direction. The magnitude of the shift is inversely proportional to the receiver bandwidth in agreement with theory. Comparison of MRI with CT showed the same image shift. We concluded that the discrepancy is caused by MR image distortion due to a difference in susceptibility effects between the phantom and the fiducial markers of the Leksell localization box.     

Diffeomorphic susceptibility artifact correction of diffusion-weighted magnetic resonance images

Diffusion-weighted magnetic resonance imaging is a key investigation technique in modern neuroscience. In clinical settings, diffusion-weighted imaging and its extension to diffusion tensor imaging (DTI) are usually performed applying the technique of echo-planar imaging (EPI). EPI is the commonly available ultrafast acquisition technique for single-shot acquisition with spatial encoding in a Cartesian system. A drawback of these sequences is their high sensitivity against small perturbations of the magnetic field, caused, e.g., by differences in magnetic susceptibility of soft tissue, bone and air. The resulting magnetic field inhomogeneities thus cause geometrical distortions and intensity modulations in diffusion-weighted images. This complicates the fusion with anatomical T1- or T2-weighted MR images obtained with conventional spin- or gradient-echo images and negligible distortion. In order to limit the degradation of diffusion-weighted MR data, we present here a variational approach based on a reference scan pair with reversed polarity of the phase- and frequency-encoding gradients and hence reversed distortion. The key novelty is a tailored nonlinear regularization functional to obtain smooth and diffeomorphic transformations. We incorporate the physical distortion model into a variational image registration framework and derive an accurate and fast correction algorithm. We evaluate the applicability of our approach to distorted DTI brain scans of six healthy volunteers. For all datasets, the automatic correction algorithm considerably reduced the image degradation. We show that, after correction, fusion with T1- or T2-weighted images can be obtained by a simple rigid registration. Furthermore, we demonstrate the improvement due to the novel regularization scheme. Most importantly, we show that it provides meaningful, i.e. diffeomorphic, geometric transformations, independent of the actual choice of the regularization parameters.     

Predictive sensor based x-ray calibration using a physical model

Many computer assisted surgery systems are based on intraoperative x-ray images. To achieve reliable and accurate results these images have to be calibrated concerning geometric distortions, which can be distinguished between constant distortions and distortions caused by magnetic fields. Instead of using an intraoperative calibration phantom that has to be visible within each image resulting in overlaying markers, the presented approach directly takes advantage of the physical background of the distortions. Based on a computed physical model of an image intensifier and a magnetic field sensor, an online compensation of distortions can be achieved without the need of an intraoperative calibration phantom. The model has to be adapted once to each specific image intensifier through calibration, which is based on an optimization algorithm systematically altering the physical model parameters, until a minimal error is reached. Once calibrated, the model is able to predict the distortions caused by the measured magnetic field vector and build an appropriate dewarping function. The time needed for model calibration is not yet optimized and takes up to 4 h on a 3 GHz CPU. In contrast, the time needed for distortion correction is less than 1 s and therefore absolutely acceptable for intraoperative use. First evaluations showed that by using the model based dewarping algorithm the distortions of an XRII with a 21 cm FOV could be significantly reduced. The model was able to predict and compensate distortions by approximately 80% to a remaining error of 0.45 mm (max) (0.19 mm rms).     

Fast acceleration-encoded magnetic resonance imaging

Direct acceleration imaging with high spatial resolution was implemented and tested. The well-known principle of phase encoding motion components was applied. Suitable gradient switching provides a signal phase shift proportional to the acceleration perpendicular to the slice in the first scan of the sequences. An additional scan serving as a reference was recorded for compensation of phase effects due to magnetic field inhomogeneities. The first scan compensated for phase shifts from undesired first- and second-order motions; the second scan was completely insensitive to velocity and acceleration in all directions. Advantages of the proposed two-step technique compared to former approaches with Fourier acceleration encoding (with several phase encoding steps) are relatively short echo times and short total measuring times. On the other hand, the new approach does not allow us to assess the velocity or acceleration spectrum simultaneously. The capabilities of the sequences were tested on a modern 1.5 T whole body MR unit providing relatively high gradient amplitudes (25 mT/m) and short rise times (600 micros to maximum amplitude). The results from a mechanical acceleration phantom showed a standard deviation of 0.3 m/s2 in sequences with an acceleration range between -12 and 12 m/s2. This range covers the expected maximum acceleration in the human aorta of 10 m/s2. Further tests were performed on a stenosis phantom with a variable volume flow rate to assess the flow characteristics and possible displacement artifacts of the sequences. Preliminary examinations of volunteers demonstrate the potential applicability of the technique in vivo.     

Characteristics and quality assurance of a dedicated open 0.23 T MRI for radiation therapy simulation

A commercially available open MRI unit is under routine use for radiation therapy simulation. The effects of a gradient distortion correction (GDC) program used to post process the images were assessed by comparison with the known geometry of a phantom. The GDC reduced the magnitude of the distortions at the periphery of the axial images from 12 mm to 2 mm horizontally along the central axis and distortions exceeding 20 mm were reduced to as little as 2 mm at the image periphery. Coronal and sagittal scans produced similar results. Coalescing these data into distortion as a function of radial distance, we found that for radial distances of <10 cm, the distortion after GDC was <2 mm and for radial distances up to 20 cm, the distortion was <5 mm. The dosimetric errors resulting from homogeneous dose calculations with this level of distortion of the external contour is <2%. A set of triangulation lasers has been added to establish a virtual isocenter for convenient setup and marking of patients and phantoms. Repeated measurements of geometric phantoms over several months showed variations in position between the virtual isocenter and the magnetic isocenter were constrained to <2 mm. Additionally, the interscan variations of 12 randomly selected points in space defined by a rectangular grid phantom was found to be within the intraobserver error of approximately 1 mm in the coronal, sagittal, and transverse planes. Thus, the open MRI has sufficient geometric accuracy for most radiation therapy planning and is temporally stable.     

Artefacts in multi-echo T2 imaging for high-precision gel dosimetry: II. Analysis of B1-field inhomogeneity

In BANG gel dosimetry, the spin-spin relaxation rate, R2 = 1/T2, is related to radiation dose that has been delivered to a gel phantom. R2 is calculated by fitting the pixel intensities of a set of differently T2-weighted base images. The accuracy that is aimed for in this quantitative MR application is about 5% relative to the maximum dose. In a conventional imaging MR scanner, however, several imaging artefacts may perturb the final dose map. These deviations manifest themselves as either a deformation of the dose map or an inaccuracy of the dose pixel value. Inaccuracies in the dose maps are caused by both spatial and temporal deviations in signal intensities during scanning. This study deals with B1-field inhomogeneities as a source of dose inaccuracy. First, the influence of B1-field inhomogeneities on slice profiles is investigated using a thin-slice phantom. Secondly, a FLASH sequence is used to map the B1-field by assessing the effective flip angle in each voxel of a homogeneous phantom. In addition, both experiments and computer simulations revealed the effects of B1 field inhomogeneities on the measured R2. This work offers a method to correct R2 maps for B1 -field inhomogeneities.     

A quantitative study of the pixel-shifting, blurring and nonlinear distortions in MRI images caused by the presence of metal implants

Magnetic resonance imaging has found an increasing number of medical applications in recent years due to its technical merits as well as its non-invasive nature. However, its full potential has been severely limited by magnetic susceptibility difference artefacts caused by the presence of ferromagnetic sources such as orthopedic implants, dental work or metallic needles used in neurosurgery. In this study, we propose a method to numerically quantify the distortions resulting from the magnetic susceptibility differences by investigating the phenomena from three perspectives: (1) pixel displacement, (2) blurring and (3) nonlinearity. For this purpose, phantom images obtained from a magnetic resonance scanner were studied. Attempts made to reconstruct an ideal image from its distorted version by appropriately compensating for the three types of distortions yielded encouraging results.     

A quantitative study of the pixel-shifting,blurring and nonlinear distortions in MRI images caused by the presence of metal implants

Magnetic resonance imaging has found an increasing number of medical applications in recent years due to its technical merits as well as its non-invasive nature. However, its full potential has been severely limited by magnetic susceptibility difference artefacts caused by the presence of ferromagnetic sources such as orthopedic implants, dental work or metallic needles used in neurosurgery. In this study, we propose a method to numerically quantify the distortions resulting from the magnetic susceptibility differences by investigating the phenomena from three perspectives: (1) pixel displacement, (2) blurring and (3) nonlinearity. For this purpose, phantom images obtained from a magnetic resonance scanner were studied. Attempts made to reconstruct an ideal image from its distorted version by appropriately compensating for the three types of distortions yielded encouraging results.     

Accuracy of device-specific 2D and 3D image distortion correction algorithms for magnetic resonance imaging of the head provided by a manufacturer

For the application of magnetic resonance imaging (MRI) in precision radiotherapy, image distortions must be reduced to a minimum to maintain geometrical accuracy. Recently, two-dimensional (2D) and three-dimensional (3D) algorithms for MRI-device-specific distortion corrections were developed by the manufacturers of MRI devices. A previously developed phantom (Karger C P et al 2003 Phys. Med. Biol. 48 211-21) was used to quantify and assess the size of geometrical image distortions before and after application of the 2D and 3D correction algorithm in the head region. Four different types of MRI devices with different gradient systems were measured. For comparison, measurements were also performed with two computed tomography (CT) devices. Mean distortions of up to 4.6+/-1.4 mm (maximum: 5.8 mm) were found prior to the correction. After the correction, the mean distortions were well below 2.0 mm in most cases. Distortions in the CT images were below or equal to 1.0 mm on average. Generally, the 3D algorithm produced comparable or better results than the 2D algorithm. The remaining distortions after the correction appear to be acceptable for fractionated radiotherapy.     

A tool for validating MRI-guided strategies: a digital breathing CT/MRI phantom of the abdominal site

Dynamic magnetic resonance imaging (MRI) is emerging as the elected image modality for organ motion quantification and management in image-guided radiotherapy. However, the lack of validation tools is an open issue for image guidance in the abdominal and thoracic organs affected by organ motion due to respiration. We therefore present an abdominal four-dimensional (4D) CT/MRI digital phantom, including the estimation of MR tissue parameters, simulation of dedicated abdominal MR sequences, modeling of radiofrequency coil response and noise, followed by k-space sampling and image reconstruction. The phantom allows the realistic simulation of images generated by MR pulse sequences with control of scan and tissue parameters, combined with co-registered CT images. In order to demonstrate the potential of the phantom in a clinical scenario, we describe the validation of a virtual T1-weighted 4D MRI strategy. Specifically, the motion extracted from a T2-weighted 4D MRI is used to warp a T1-weighted breath-hold acquisition, with the aim of overcoming trade-offs that limit T1-weighted acquisitions. Such an application shows the applicability of the digital CT/MRI phantom as a validation tool, which should be especially useful for cases unsuited to obtain real imaging data.     

Improved target volume characterization in stereotactic treatment planning of brain lesions by using high-resolution BOLD MR-venography.

In this methodological paper I report the stereotactic correlation of different magnetic resonance imaging (MRI) techniques [MR angiography (MRA), MRI, blood bolus tagging (STAR), functional MRI, and high-resolution BOLD venography (HRBV)] in patients with cerebral arterio-venous malformations (AVM) and brain tumors. The patient's head was fixed in a stereotactic localization system which is usable in both MR-systems and linear accelerator installations. Using phantom measurements global geometric MR image distortions can be 'corrected' (reducing displacements to the size of a pixel) by calculations based on modeling the distortion as a fourth-order two-dimensional polynomial. Further object-induced local distortions can be corrected by additionally measured field maps. Using this method multimodality matching could be performed automatically as long as all images are acquired in the same examination and the patient is sufficiently immobilized to allow precise definition of the target volume. Information about the hemodynamics of the AVM was provided by a dynamic MRA with the STAR technique, leading to an improved definition of the size of the nidus, the origin of the feeding arteries, whereas HRBV imaging yielded detailed and improved information about the venous pattern and drainage. In addition, functional MRI was performed in patients with lesions close to the primary motor cortex area, leading to an improved definition of structures at risk for the high-dose application in radiosurgery. In patients with brain tumors the potential of HRBV to probe tumor angiogenesis and its use in intensity-modulated treatment planning is still hampered by the open question of how to translate a BOLD signal pattern measured in the tumor to a dose distribution, which should be addressed in future studies.     

Heterogeneity phantoms for visualization of 3D dose distributions by MRI-based polymer gel dosimetry

Heterogeneity corrections in dose calculations are necessary for radiation therapy treatment plans. Dosimetric measurements of the heterogeneity effects are hampered if the detectors are large and their radiological characteristics are not equivalent to water. Gel dosimetry can solve these problems. Furthermore, it provides three-dimensional (3D) dose distributions. We used a cylindrical phantom filled with BANG-3 polymer gel to measure 3D dose distributions in heterogeneous media. The phantom has a cavity, in which water-equivalent or bone-like solid blocks can be inserted. The irradiated phantom was scanned with an magnetic resonance imaging (MRI) scanner. Dose distributions were obtained by calibrating the polymer gel for a relationship between the absorbed dose and the spin-spin relaxation rate of the magnetic resistance (MR) signal. To study dose distributions we had to analyze MR imaging artifacts. This was done in three ways: comparison of a measured dose distribution in a simulated homogeneous phantom with a reference dose distribution, comparison of a sagittally scanned image with a sagittal image reconstructed from axially scanned data, and coregistration of MR and computed-tomography images. We found that the MRI artifacts cause a geometrical distortion of less than 2 mm and less than 10% change in the dose around solid inserts. With these limitations in mind we could make some qualitative measurements. Particularly we observed clear differences between the measured dose distributions around an air-gap and around bone-like material for a 6 MV photon beam. In conclusion, the gel dosimetry has the potential to qualitatively characterize the dose distributions near heterogeneities in 3D.     

Mapping of the prostate in endorectal coil-based MRI/MRSI and CT: a deformable registration and validation study

The endorectal coil is being increasingly used in magnetic resonance imaging (MRI) and MR spectroscopic imaging (MRSI) to obtain anatomic and metabolic images of the prostate with high signal-to-noise ratio (SNR). In practice, however, the use of endorectal probe inevitably distorts the prostate and other soft tissue organs, making the analysis and the use of the acquired image data in treatment planning difficult. The purpose of this work is to develop a deformable image registration algorithm to map the MRI/MRSI information obtained using an endorectal probe onto CT images and to verify the accuracy of the registration by phantom and patient studies. A mapping procedure involved using a thin plate spline (TPS) transformation was implemented to establish voxel-to-voxel correspondence between a reference image and a floating image with deformation. An elastic phantom with a number of implanted fiducial markers was designed for the validation of the quality of the registration. Radiographic images of the phantom were obtained before and after a series of intentionally introduced distortions. After mapping the distorted phantom to the original one, the displacements of the implanted markers were measured with respect to their ideal positions and the mean error was calculated. In patient studies, CT images of three prostate patients were acquired, followed by 3 Tesla (3 T) MR images with a rigid endorectal coil. Registration quality was estimated by the centroid position displacement and image coincidence index (CI). Phantom and patient studies show that TPS-based registration has achieved significantly higher accuracy than the previously reported method based on a rigid-body transformation and scaling. The technique should be useful to map the MR spectroscopic dataset acquired with ER probe onto the treatment planning CT dataset to guide radiotherapy planning.