Coregistration Errors Research Articles

Clinical Application of In-Room Positron Emission Tomography for In Vivo Treatment Monitoring in Proton Radiation Therapy

The purpose of this study is to evaluate the potential of using in-room positron emission tomography (PET) for treatment verification in proton therapy and for deriving suitable PET scan times. Nine patients undergoing passive scattering proton therapy underwent scanning immediately after treatment with an in-room PET scanner. The scanner was positioned next to the treatment head after treatment. The Monte Carlo (MC) method was used to reproduce PET activities for each patient. To assess the proton beam range uncertainty, we designed a novel concept in which the measured PET activity surface distal to the target at the end of range was compared with MC predictions. The repositioning of patients for the PET scan took, on average, approximately 2 minutes. The PET images were reconstructed considering varying scan times to test the scan time dependency of the method. The measured PET images show overall good spatial correlations with MC predictions. Some discrepancies could be attributed to uncertainties in the local elemental composition and biological washout. For 8 patients treated with a single field, the average range differences between PET measurements and computed tomography (CT) image-based MC results were <5 mm (<3 mm for 6 of 8 patients) and root-mean-square deviations were 4 to 11 mm with PET-CT image co-registration errors of approximately 2 mm. Our results also show that a short-length PET scan of 5 minutes can yield results similar to those of a 20-minute PET scan. Our first clinical trials in 9 patients using an in-room PET system demonstrated its potential for invivo treatment monitoring in proton therapy. For a quantitative range prediction with arbitrary shape of target volume, we suggest using the distal PET activity surface.

Single subject fMRI test–retest reliability metrics and confounding factors

While the fMRI test-retest reliability has been mainly investigated from the point of view of group level studies, here we present analyses and results for single-subject test-retest reliability. One important aspect of group level reliability is that not only does it depend on between-session variance (test-retest), but also on between-subject variance. This has partly led to a debate regarding which reliability metric to use and how different sources of noise contribute to between-session variance. Focusing on single subject reliability allows considering between-session only. In this study, we measured test-retest reliability in four behavioural tasks (motor mapping, covert verb generation, overt word repetition, and a landmark identification task) to ensure generalisation of the results and at three levels of data processing (time-series correlation, t value variance, and overlap of thresholded maps) to understand how each step influences the other and how confounding factors influence reliability at each of these steps. The contributions of confounding factors (scanner noise, subject motion, and coregistration) were investigated using multiple regression and relative importance analyses at each step. Finally, to achieve a fuller picture of what constitutes a reliable task, we introduced a bootstrap technique of within- vs. between-subject variance. Our results show that (i) scanner noise and coregistration errors have little contribution to between-session variance (ii) subject motion (especially correlated with the stimuli) can have detrimental effects on reliability (iii) different tasks lead to different reliability results. This suggests that between-session variance in fMRI is mostly caused by the variability of underlying cognitive processes and motion correlated with the stimuli rather than technical limitations of data processing.

Photo-guided sentinel node mapping in breast cancer using marker-free photo-gamma fusion lymphoscintigraphy.

Photo-gamma fusion lymphoscintigraphy (PGFLS) was developed by overlying a conventional planar gamma image on a photograph for the guidance of sentinel node biopsy. The feasibility and accuracy of PGFLS was assessed in breast cancer patients. A digital camera and a gamma camera were coordinated to obtain photograph and gamma images from the same angle. Using the distance to the object and calibration acquisition with a flat phantom and radioactive markers, PGFLS was performed both in phantom and in patients without fiducial markers. Marker-free PGFLS was verified using flat phantom, anthropomorphic phantom with markers simulating sentinel nodes and breast cancer patients. In addition, the depth of the radioactive marker or sentinel node was calculated using two gamma images taken at right angles. The feasibility and accuracy of PGFLS were assessed in terms of mismatch errors of co-registration and depth with reference to the data from SPECT/CT. The mismatch error was less than 6mm in the flat phantom image at a distance from 50 to 62cm without misalignment. In the anthropomorphic phantom study, co-registration error was 0.42 ± 0.29cm; depth error was 0.51 ± 0.37cm, which was well correlated with the reference value on SPECT/CT (x scale: R(2) = 0.99, p < 0.01; y scale: R(2) = 0.99, p < 0.01; depth: R(2) = 0.99, p < 0.01). In ten patients with breast cancer referred for lympho-SPECT/CT, PGFSL enabled photo-guided sentinel lymph node mapping with acceptable accuracy (co-registration error, 0.47 ± 0.24cm; depth error, 1.20 ±0.41cm). The results from PGFSL showed close correlation with those from SPECT/CT (x scale: R(2) = 0.99, p < 0.01; y scale: R(2) = 0.98, p < 0.01; depth: R(2) = 0.77, p < 0.01). The novel and convenient PGFLS technique is clinically feasible, showing acceptable accuracy and providing additional visual and quantitative information for sentinel lymph node mapping. This approach will facilitate photo-guided sentinel lymph node dissection in breast cancer.

Commissioning of MRI‐only based treatment planning procedure for external beam radiotherapy of prostate

In radiotherapy, target tissues are defined best on MR images due to their superior soft tissue contrast. Computed tomography imaging is geometrically accurate and it is needed for dose calculation and generation of reference images for treatment localization. Co-registration errors between MR and computed tomography images can be eliminated using magnetic resonance imaging-only based treatment planning. Use of ionizing radiation can be avoided which is especially important in adaptive treatments requiring several re-scans. We commissioned magnetic resonance imaging-only based procedure for external radiotherapy, treatment planning of the prostate cancer. Geometrical issues relevant in radiotherapy, were investigated including quality assurance testing of the scanner, evaluation of the displacement of skin contour and radiosensitive rectum wall, and detection of intraprostatic fiducial gold seed markers used for treatment localization. Quantitative analysis was carried out for 30 randomly chosen patients. Systematic geometrical errors were within 2.2 mm. The gold seed markers were correctly identified for 29 out of the 30 patients. Positions of the seed midpoints were consistent within 1.3 mm in magnetic resonance imaging and computed tomography. Positional error of rectal anterior wall due to susceptibility effect was minimal. Geometrical accuracy of the investigated equipment and procedure was sufficient for magnetic resonance imaging-only based radiotherapy, treatment planning of the prostate cancer including treatment virtual simulation.

Hybrid ultra‐low‐field MRI and magnetoencephalography system based on a commercial whole‐head neuromagnetometer

Ultra-low-field MRI uses microtesla fields for signal encoding and sensitive superconducting quantum interference devices for signal detection. Similarly, modern magnetoencephalography (MEG) systems use arrays comprising hundreds of superconducting quantum interference device channels to measure the magnetic field generated by neuronal activity. In this article, hybrid MEG-MRI instrumentation based on a commercial whole-head MEG device is described. The combination of ultra-low-field MRI and MEG in a single device is expected to significantly reduce coregistration errors between the two modalities, to simplify MEG analysis, and to improve MEG localization accuracy. The sensor solutions, MRI coils (including a superconducting polarizing coil), an optimized pulse sequence, and a reconstruction method suitable for hybrid MEG-MRI measurements are described. The performance of the device is demonstrated by presenting ultra-low-field-MR images and MEG recordings that are compared with data obtained with a 3T scanner and a commercial MEG device.

T1/T2*-weighted MRI provides clinically relevant pseudo-CT density data for the pelvic bones in MRI-only based radiotherapy treatment planning

Background and purpose. In radiotherapy (RT), target soft tissues are best defined on MR images. In several cases, CT imaging is needed only for dose calculation and generation of digitally reconstructed radiographs (DRRs). Image co-registration errors between MRI and CT can be avoided by using MRI-only based treatment planning, especially in the pelvis. Since electron density information can not be directly derived from the MRI, a method is needed to convert MRI data into CT like data. We investigated whether there is a relationship between MRI intensity and Hounsfield unit (HU) values for the pelvic bones. The aim was to generate a method to convert bone MRI intensity into HU data surrogate for RT treatment planning. Material and methods. The HU conversion model was generated for 10 randomly chosen prostate cancer patients and independent validation was performed in another 10 patients. Data consisted of 800 image voxels chosen within the pelvic bones in both T1/T2*-weighted gradient echo and CT images. Relation between MRI intensity and electron density was derived from calibrated HU-values. The proposed method was tested by constructing five “pseudo”-CT series. Results. We found that the MRI intensity is related to the HU value within a HU range from 0 to 1400 within the pelvic bones. The mean prediction error of the conversion model was 135 HU. Dose calculation based on the pseudo-CT images was accurate and the generated DRRs were of good quality. Conclusions. The proposed method enables generation of clinically relevant pseudo-CT data for the pelvic bones from one MRI series. It is simpler than previously reported approaches which require either acquisition of several MRI series or T2* maps with special imaging sequences. The method can be applied with commercial clinical image processing software. The application requires segmentation of the bones in the MR images.

An in-depth simulation of EnMAP acquisition geometry

The future hyperspectral satellite EnMAP (Environmental Mapping and Analysis Program) uses two separate sensors for the acquisition of VNIR and SWIR imagery. Due to their geometric configuration, the SWIR and VNIR instruments map the same positions on the ground with a time delay of 88ms. Coupled with attitude controller inaccuracies this leads to an estimated co-registration error between SWIR and VNIR higher than the maximum 0.2 pixels designated in the specifications of EnMAP imagery. It is assumed that, by approximating or interpolating the real attitude and geometrically correcting the images, this co-registration error can be significantly reduced. To validate these assumptions, a geometric simulator was developed at the German Aerospace Center DLR which is responsible for the development of the ground segment of EnMAP. The implemented simulator, together with an evaluation of the absolute and relative accuracy, performed using this simulator, are presented in this article. The obtained results demonstrate that the desired co-registration accuracy between SWIR and VNIR imagery can be achieved by using Spline or Chebyshev approximation for the attitude reconstruction but not by using Lagrange interpolation.

Usefulness and pitfalls of MAA SPECT/CT in identifying digestive extrahepatic uptake when planning liver radioembolization

PurposeIdentifying gastroduodenal uptake of 99mTc-macroaggregated albumin (MAA), which is associated with an increased risk of ulcer disease, is a crucial part of the therapeutic management of patients undergoing radioembolization for liver tumours. Given this context, the use of MAA single photon emission computed tomography (SPECT)/CT may be essential, but the procedure has still not been thoroughly evaluated. The aim of this retrospective study was to determine the effectiveness of MAA SPECT/CT in identifying digestive extrahepatic uptake, while determining potential diagnostic pitfalls.MethodsOverall, 139 MAA SPECT/CT scans were performed on 103 patients with different hepatic tumour types. Patients were followed up for at least 6 months according to standard requirements.ResultsDigestive, or digestive-like, uptake other than free pertechnetate was identified in 5.7% of cases using planar imaging and in 36.6% of cases using SPECT/CT. Uptake sites identified by SPECT/CT included the gastroduodenal region (3.6%), gall bladder (12.2%), portal vein thrombosis (6.5%), hepatic artery (6.5%), coil embolization site (2.1%) as well as falciform artery (5.0%). For 2.1% of explorations, a coregistration error between SPECT and CT imaging could have led to a false diagnosis by erroneously attributing an uptake site to the stomach or gall bladder, when the uptake actually occurred in the liver.ConclusionSPECT/CT is more efficacious than planar imaging in identifying digestive extrahepatic uptake sites, with extrahepatic uptake observed in one third of scans using the former procedure. However, more than half of the uptake sites in our study were vascular in nature, without therapeutic implications. The risk of coregistration errors must also be kept in mind.

An upper-bound metric for characterizing spectral and spatial coregistration errors in spectral imaging

Coregistration errors in multi- and hyperspectral imaging sensors arise when the spatial sensitivity pattern differs between bands or when the spectral response varies across the field of view, potentially leading to large errors in the recorded image data. In imaging spectrometers, spectral and spatial offset errors are customarily specified as "smile" and "keystone" distortions. However these characteristics do not account for errors resulting from variations in point spread function shape or spectral bandwidth. This paper proposes improved metrics for coregistration error both in the spatial and spectral dimensions. The metrics are essentially the integrated difference between point spread functions. It is shown that these metrics correspond to an upper bound on the error in image data. The metrics enable estimation of actual data errors for a given image, and can be used as part of the merit function in optical design optimization, as well as for benchmarking of spectral image sensors.

Comparison of MISR and Meteosat-9 cloud-motion vectors

[1] Stereo motion vectors (SMVs) from the Multiangle Imaging SpectroRadiometer (MISR) were evaluated against Meteosat-9 cloud-motion vectors (CMVs) over a one-year period. In general, SMVs had weaker westerlies and southerlies than CMVs at all latitudes and levels. The E-W wind comparison showed small vertical variations with a mean difference of −0.4 m s−1, −1 m s−1, −0.7 m s−1 and corresponding rmsd of 2.4 m s−1, 3.8 m s−1, 3.5 m s−1for low-, mid-, and high-level clouds, respectively. The N-S wind discrepancies were larger and steadily increased with altitude, having a mean difference of −0.8 m s−1, −2.9 m s−1, −4.4 m s−1 and rmsd of 3.5 m s−1, 6.9 m s−1, 9.5 m s−1at low, mid, and high levels. The best overall agreement was found in marine stratocumulus off Namibia, while differences were larger in the Tropics and convective clouds. The SMVs were typically assigned to higher altitudes than CMVs. Attributing each observed height difference to MISR and/or Meteosat-9 retrieval biases will require further research; nevertheless, we already identified a few regions and cloud types where CMV height assignment seemed to be the one in error. In thin mid- and high-level clouds over Africa and Arabia as well as in broken marine boundary layer clouds the 10.8-μm brightness temperature-based heights were often biased low due to radiance contributions from the warm surface. Contrarily, low-level CMVs in the South Atlantic were frequently assigned to mid levels by the CO2-slicing method in multilayer situations. We also noticed an apparent cross-swath dependence in SMVs, whereby retrievals were less accurate on the eastern side of the MISR swath than on the western side. This artifact was traced back to sub-pixel MISR co-registration errors, which introduced cross-swath biases in E-W wind, N-S wind, and height of 0.6 m s−1, 2.6 m s−1, and 210 m.

Automated Least-Squares Calibration of the Coregistration Parameters for a Micro PET-CT System

PET-CT coregistration parameters can be derived from PET and CT images of a four-point-source calibration phantom for a micro PET-CT scanner. An automated segmentation method has been developed, based on thresholding and application of constraints on the sizes of point sources in the images. After point sources are identified on PET and CT images, coregistration is performed using an analytic rigid-body registration algorithm which is based on singular value decomposition and minimization of the coregistration error. The coregistration parameters thus derived can then be applied to coregister other PET and CT images from the same system. Twenty PET-CT images of the calibration phantom at various locations and/or orientations were obtained on a Siemens Inveon® Multi-Modality scanner. We tested the use of from 1 to 10 data sets to derive the coregistration parameters, and found that the coregistration accuracy improves with increasing number of data sets until it stabilizes. Coregistration of PET-CT images with an accuracy of 0.33±0.11 mm has been achieved by this method on the Inveon Multi-Modality scanner.

Geo-referencing remote images for precision agriculture using artificial terrestrial targets

The aim of this paper is to assess co-registration errors in remote imagery through the AUGEO system, which consists of geo-referenced coloured tarps acting as terrestrial targets (TT), captured in the imagery and semi-automatically recognised by AUGEO2.0® software. This works as an add-on of ENVI® for image co-registration. To validate AUGEO, TT were placed in the ground, and remote images from satellite Quick Bird (QB), airplanes and unmanned aerial vehicles (UAV) were taken at several locations in Andalusia (southern Spain) in 2008 and 2009. Any geo-referencing system tested showed some error in comparison with the Differential Global Positioning System (DGPS)-geo-referenced verification targets. Generally, the AUGEO system provided higher geo-referencing accuracy than the other systems tried. The root mean square errors (RMSE) from the panchromatic and multi-spectral QB images were around 8 and 9 m, respectively and, once co-registered by AUGEO, they were about 1.5 and 2.5 m, for the same images. Overlapping the QB-AUGEO-geo-referenced image and the National Geographic Information System (NGIS) produced a RMSE of 6.5 m, which is hardly acceptable for precision agriculture. The AUGEO system efficiently geo-referenced farm airborne images with a mean accuracy of about 0.5–1.5 m, and the UAV images showed a mean accuracy of 1.0–4.0 m. The geo-referencing accuracy of an image refers to its consistency despite changes in its spatial resolution. A higher number of TT used in the geo-referencing process leads to a lower obtained RMSE. For example, for an image of 80 ha, about 10 and 17 TT were needed to get a RMSE less than about 2 and 1 m. Similarly, with the same number of TT, accuracy was higher for smaller plots as compared to larger plots. Precision agriculture requires high spatial resolution images (i.e., <1.5 m pixel−1), accurately geo-referenced (errors <1–2 m). With the current DGPS technology, satellite and airplane images hardly meet this geo-referencing requirement; consequently, additional co-registration effort is needed. This can be achieved using geo-referenced TT and AUGEO, mainly in areas where no notable hard points are available.

Image Autocoregistration and Interferogram Estimation Using Extended COMET-EXIP Method

In this paper, an extended COvariance Matching Estimation Techniques-Extend Invariance Principle (COMET-EXIP) method is proposed to estimate interferometric synthetic aperture radar or interferometric synthetic aperture sonar (InSAS) interferometric phase in the presence of large coregistration errors, even up to one pixel. First, the extended COMET-EXIP method is presented for the application of joint-pixel-model-based interferogram estimation, through choosing a novel “unstructured model” in terms of the parameters to be estimated and decoupling the interesting parameters from the uninteresting “nuisance parameters.” Then, a fast algorithm of COMET-EXIP is proposed for the interferometric phase estimation. Finally, the ambiguity problem of the COMET-EXIP method is solved without introducing performance degradation. The simulated data and real data from the trial InSAS and X-SAR are used to verify the validity of the method. The results show that the method is robust for a wide range of signal-to-noise ratio and has a good performance on both fringe preserving and noise suppressing. In addition, the same computational speed level of the proposed method as that of the pivoting mean filtering is very attractive.

Simulated impact of sample plot size and co-registration error on the accuracy and uncertainty of LiDAR-derived estimates of forest stand biomass

Regression has been widely applied in Light Detection And Ranging (LiDAR) remote sensing to spatially extend predictions of total aboveground biomass (TAGB) and other biophysical properties over large forested areas. Sample (field) plot size has long been considered a key sampling design parameter and focal point for optimization in forest surveys, because of its impact on sampling effort and the estimation accuracy of forest inventory attributes. In this study, we demonstrate how plot size and co-registration error interact to influence the estimation of LiDAR canopy height and density metrics, regression model coefficients, and the prediction accuracy of least-squares estimators of TAGB. We made use of simulated forest canopies and synthetic LiDAR point clouds, so that we could maintain strict control over the spatial scale and complexity of forest scenes, as well as the magnitude and type of planimetric error inherent in ground-reference and LiDAR datasets. Our results showed that predictions of TAGB improved markedly as plot size increased from 314 (10 m radius) to 1964 m 2 (25 m radius). The co-registration error (spatial overlap) between ground-reference and LiDAR samples negatively impacted the estimation of LiDAR metrics, regression model fit, and the prediction accuracy of TAGB. We found that larger plots maintained a higher degree of spatial overlap between ground-reference and LiDAR datasets for any given GPS error, and were therefore more resilient to the ill effects of co-registration error compared to small plots. The impact of co-registration error was more pronounced in tall, spatially heterogeneous stands than short, homogeneous stands. We identify and briefly discuss three possible ways that LiDAR data could be used to optimize plot size, sample selection, and the deployment of GPS resources in forest biomass surveys.

Ground settlement monitoring based on temporarily coherent points between two SAR acquisitions

An InSAR analysis approach for identifying and extracting the temporarily coherent points (TCP) that exist between two SAR acquisitions and for determining motions of the TCP is presented for applications such as ground settlement monitoring. TCP are identified based on the spatial characteristics of the range and azimuth offsets of coherent radar scatterers. A method for coregistering TCP based on the offsets of TCP is given to reduce the coregistration errors at TCP. An improved phase unwrapping method based on the minimum cost flow (MCF) algorithm and local Delaunay triangulation is also proposed for sparse TCP data. The proposed algorithms are validated using a test site in Hong Kong. The test results show that the algorithms work satisfactorily for various ground features.

Multi-Channel SAR-GMTI Method Robust to Coregistration Error of SAR Images

High accuracy in complex images coregistration is essential in ground moving target indication (GMTI) processing of synthetic aperture radar (SAR) data, normally termed SAR-GMTI. The clutter suppression performance is proportional to the coregistration accuracy. In this correspondence, we propose a new SAR-GMTI approach, which is robust to the SAR images coregistration error. The observed clutter-plus-noise vector is built using the neighboring pixels as the first step, thus the corresponding covariance matrix, termed as joint covariance matrix, can be estimated. The joint noise subspace is obtained by eigen-decomposing of the joint covariance matrix. The clutter is suppressed successfully by a projecting operation, i.e., a projection of the joint observed vector into the resulting joint noise subspace. The information of neighboring pixels is substantially exploited in the clutter suppression, resulting in the robustness, even at the presence of large coregistration error. Both the simulated and real multi-channel airborne data are used to validate the proposed approach.

On robust beamforming for InSAR phase estimation

In InSAR signal processing, the performance of the interferometric phase estimation degrades in the case of image coregistration errors and spatial steering vector errors. In this letter, a method of interferometric phase estimation based on robust beamforming is introduced. The proposed method can take advantage of the optimal weighted joint data (OWJD) vector and of the robust adaptive beamformer to auto-coregister the SAR images and mitigate the effect of the spatial steering error on the interferometric phase estimation. Theoretical analysis and experimental results show that the method can provide accurate estimation of the interferometric phase in the presence of the coregistration error and spatial steering vector error. Numerical results on the simulated and real data demonstrate the effectiveness and robustness of the proposed method.

Estimation Method for InSAR Interferometric Phase Based on Generalized Correlation Steering Vector

We propose a new method, based on the generalized correlation steering vector, to estimate synthetic aperture radar interferometry (InSAR) interferometric phase. In this method the generalized correlation steering vector with a large coregistration error is determined according to the joint data vector. The generalized correlation steering vector is then used to estimate the InSAR interferometric phase. The method can simultaneously auto-coregister the synthetic aperture radar (SAR) images and reduce the interferometric phase noise. Theoretical analysis and computer simulation results show that the method can provide an accurate estimate of the terrain interferometric phase (interferogram), even if the coregistration error reaches one pixel.

Perito Moreno Glacier (Argentina) flow estimation by COSMO SkyMed sequence of high-resolution SAR-X imagery

The Perito Moreno Glacier (Los Glaciares National Park, Patagonia, Argentina) ice stream has been investigated using X-band SAR amplitude images collected from February to December 2009 (excluding June) at time intervals of 8 and 16 days by the COSMO SkyMed satellites. The high spatial resolution dataset (10 × 10 km swath, 1 m per pixel) reveals many details of the glacier surface and the ice velocity fields extracted from sequential image pairs by the maximum cross-correlation method cover most of the glacier area. The efficiency of the cross-correlation method improves with high spatial resolution imagery and the co-registration error of the image pair also reduces. The glacier motion detected in the period shows variability on the time scale considered, evidencing the areas of higher dynamics under different speed regimes.

An Estimation Method for InSAR Interferometric Phase Based on MMSE Criterion

In this paper, we propose a method based on minimum mean squared error (mmse) criterion to estimate synthetic aperture radar interferometry (InSAR) interferometric phase. In this method, the cross-correlation coefficient vector with large coregistration error is given first, and then, the cost function under the mmse criterion is used to estimate the InSAR interferometric phase. The method can auto-coregister the SAR images and reduce the interferometric phase noise simultaneously. Theoretical analysis and computer simulation results show that the method can provide an accurate estimate of the terrain interferometric phase (interferogram) even when the coregistration error reaches 1 pixel.