Symmetric inverse consistent nonlinear registration driven by mutual information

Symmetric pyramid network for medical image inverse consistent diffeomorphic registration

Over the past few years, deep learning-based image registration methods have achieved remarkable performance in medical image analysis. However, many existing methods struggle to ensure accurate registration while preserving the desired diffeomorphic properties and inverse consistency of the final deformation field. To address the problem, this paper presents a novel symmetric pyramid network for medical image inverse consistent diffeomorphic registration. Specifically, we first encode the multi-scale images to the feature pyramids via a shared-weights encoder network and then progressively conduct the feature-level diffeomorphic registration. The feature-level registration is implemented symmetrically to ensure inverse consistency. We independently carry out the forward and backward feature-level registration and average the estimated bidirectional velocity fields for more robust estimation. Finally, we employ symmetric multi-scale similarity loss to train the network. Experimental results on three public datasets, including Mindboggle101, CANDI, and OAI, show that our method significantly outperforms others, demonstrating that the proposed network can achieve accurate alignment and generate the deformation fields with expected properties. Our code will be available at https://github.com/zhangliutong/SPnet.

INSPIRE: Intensity and spatial information-based deformable image registration

We present INSPIRE, a top-performing general-purpose method for deformable image registration. INSPIRE brings distance measures which combine intensity and spatial information into an elastic B-splines-based transformation model and incorporates an inverse inconsistency penalization supporting symmetric registration performance. We introduce several theoretical and algorithmic solutions which provide high computational efficiency and thereby applicability of the proposed framework in a wide range of real scenarios. We show that INSPIRE delivers highly accurate, as well as stable and robust registration results. We evaluate the method on a 2D dataset created from retinal images, characterized by presence of networks of thin structures. Here INSPIRE exhibits excellent performance, substantially outperforming the widely used reference methods. We also evaluate INSPIRE on the Fundus Image Registration Dataset (FIRE), which consists of 134 pairs of separately acquired retinal images. INSPIRE exhibits excellent performance on the FIRE dataset, substantially outperforming several domain-specific methods. We also evaluate the method on four benchmark datasets of 3D magnetic resonance images of brains, for a total of 2088 pairwise registrations. A comparison with 17 other state-of-the-art methods reveals that INSPIRE provides the best overall performance. Code is available at github.com/MIDA-group/inspire.

Quantifying deep grey matter atrophy using automated segmentation approaches: A systematic review of structural MRI studies

The deep grey matter (DGM) nuclei of the brain play a crucial role in learning, behaviour, cognition, movement and memory. Although automated segmentation strategies can provide insight into the impact of multiple neurological conditions affecting these structures, such as Multiple Sclerosis (MS), Huntington's disease (HD), Alzheimer's disease (AD), Parkinson's disease (PD) and Cerebral Palsy (CP), there are a number of technical challenges limiting an accurate automated segmentation of the DGM. Namely, the insufficient contrast of T1 sequences to completely identify the boundaries of these structures, as well as the presence of iso-intense white matter lesions or extensive tissue loss caused by brain injury. Therefore in this systematic review, 269 eligible studies were analysed and compared to determine the optimal approaches for addressing these technical challenges. The automated approaches used among the reviewed studies fall into three broad categories, atlas-based approaches focusing on the accurate alignment of atlas priors, algorithmic approaches which utilise intensity information to a greater extent, and learning-based approaches that require an annotated training set. Studies that utilise freely available software packages such as FIRST, FreeSurfer and LesionTOADS were also eligible, and their performance compared. Overall, deep learning approaches achieved the best overall performance, however these strategies are currently hampered by the lack of large-scale annotated data. Improving model generalisability to new datasets could be achieved in future studies with data augmentation and transfer learning. Multi-atlas approaches provided the second-best performance overall, and may be utilised to construct a "silver standard" annotated training set for deep learning. To address the technical challenges, providing robustness to injury can be improved by using multiple channels, highly elastic diffeomorphic transformations such as LDDMM, and by following atlas-based approaches with an intensity driven refinement of the segmentation, which has been done with the Expectation Maximisation (EM) and level sets methods. Accounting for potential lesions should be achieved with a separate lesion segmentation approach, as in LesionTOADS. Finally, to address the issue of limited contrast, R2*, T2* and QSM sequences could be used to better highlight the DGM due to its higher iron content. Future studies could look to additionally acquire these sequences by retaining the phase information from standard structural scans, or alternatively acquiring these sequences for only a training set, allowing models to learn the "improved" segmentation from T1-sequences alone.

Myelinating Proteins in MS Are Linked to Volumetric Brain MRI Changes

BACKGROUND AND PURPOSE

There is evidence of a relationship between promyelinating proteins and clinical multiple sclerosis (MS) activity during clinical relapse or recovery from clinical relapses. We examined the linkage between promyelinating biomarkers and volumetric changes in MS subjects during serial magnetic resonance imaging (MRI).

METHODS

We enrolled 13 MS subjects with active brain MRI scans not on disease modifying therapies. Subjects underwent baseline MRI, serum, and cerebrospinal fluid (CSF) sampling. Qualitative changes, new/resolving gadolinium, new/enlarging/diminishing T2 and T1 hypointense lesions, were compared to baseline in subsequent MRI scans, and volumetric analysis was calculated. Analysis of biomarkers on serial CSF samples was performed only in subjects with qualitative (and quantitative) changes on MRI. The study was performed at a MS Center of Excellence academic medical center.

RESULTS

There was increased CSF neural cell adhesion molecule (N-CAM) during increased qualitative T1 activity. A positive correlation between CSF and serum N-CAM and T1 lesion volume was observed. A negative correlation between serum brain-derived neurotrophic factor (BDNF) and BPH (T1 vol/T2 vol + T1 vol) was observed.

CONCLUSIONS

Increased N-CAM levels may be related to repair or remyelination following injury to the brain as shown by increased T1 volumes. Our data suggest an early kind of blood signaling that induces release of peripheral BDNF levels.

A graphical approach to optimizing variable-kernel smoothing parameters for improved deformable registration of CT and cone beam CT images

A systematic method is presented for determining optimal parameters in variable-kernel deformable image registration of cone beam CT and CT images, in order to improve accuracy and convergence for potential use in online adaptive radiotherapy. Assessed conditions included the noise constant (symmetric force demons), the kernel reduction rate, the kernel reduction percentage, and the kernel adjustment criteria. Four such parameters were tested in conjunction with reductions of 5, 10, 15, 20, 30, and 40%. Noise constants ranged from 1.0 to 1.9 for pelvic images in ten prostate cancer patients. A total of 516 tests were performed and assessed using the structural similarity index. Registration accuracy was plotted as a function of iteration number and a least-squares regression line was calculated, which implied an average improvement of 0.0236% per iteration. This baseline was used to determine if a given set of parameters under- or over-performed. The most accurate parameters within this range were applied to contoured images. The mean Dice similarity coefficient was calculated for bladder, prostate, and rectum with mean values of 98.26%, 97.58%, and 96.73%, respectively; corresponding to improvements of 2.3%, 9.8%, and 1.2% over previously reported values for the same organ contours. This graphical approach to registration analysis could aid in determining optimal parameters for Demons-based algorithms. It also establishes expectation values for convergence rates and could serve as an indicator of non-physical warping, which often occurred in cases  >0.6% from the regression line.

Inverse consistent non-rigid image registration based on robust point set matching

Background

Robust point matching (RPM) has been extensively used in non-rigid registration of images to robustly register two sets of image points. However, except for the location at control points, RPM cannot estimate the consistent correspondence between two images because RPM is a unidirectional image matching approach. Therefore, it is an important issue to make an improvement in image registration based on RPM.

Methods

In our work, a consistent image registration approach based on the point sets matching is proposed to incorporate the property of inverse consistency and improve registration accuracy. Instead of only estimating the forward transformation between the source point sets and the target point sets in state-of-the-art RPM algorithms, the forward and backward transformations between two point sets are estimated concurrently in our algorithm. The inverse consistency constraints are introduced to the cost function of RPM and the fuzzy correspondences between two point sets are estimated based on both the forward and backward transformations simultaneously. A modified consistent landmark thin-plate spline registration is discussed in detail to find the forward and backward transformations during the optimization of RPM. The similarity of image content is also incorporated into point matching in order to improve image matching.

Results

Synthetic data sets, medical images are employed to demonstrate and validate the performance of our approach. The inverse consistent errors of our algorithm are smaller than RPM. Especially, the topology of transformations is preserved well for our algorithm for the large deformation between point sets. Moreover, the distance errors of our algorithm are similar to that of RPM, and they maintain a downward trend as whole, which demonstrates the convergence of our algorithm. The registration errors for image registrations are evaluated also. Again, our algorithm achieves the lower registration errors in same iteration number. The determinant of the Jacobian matrix of the deformation field is used to analyse the smoothness of the forward and backward transformations. The forward and backward transformations estimated by our algorithm are smooth for small deformation. For registration of lung slices and individual brain slices, large or small determinant of the Jacobian matrix of the deformation fields are observed.

Conclusions

Results indicate the improvement of the proposed algorithm in bi-directional image registration and the decrease of the inverse consistent errors of the forward and the reverse transformations between two images.

Avoiding symmetry-breaking spatial non-uniformity in deformable image registration via a quasi-volume-preserving constraint

The choice of a reference image typically influences the results of deformable image registration, thereby making it asymmetric. This is a consequence of a spatially non-uniform weighting in the cost function integral that leads to general registration inaccuracy. The inhomogeneous integral measure--which is the local volume change in the transformation, thus varying through the course of the registration--causes image regions to contribute differently to the objective function. More importantly, the optimization algorithm is allowed to minimize the cost function by manipulating the volume change, instead of aligning the images. The approaches that restore symmetry to deformable registration successfully achieve inverse-consistency, but do not eliminate the regional bias that is the source of the error. In this work, we address the root of the problem: the non-uniformity of the cost function integral. We introduce a new quasi-volume-preserving constraint that allows for volume change only in areas with well-matching image intensities, and show that such a constraint puts a bound on the error arising from spatial non-uniformity. We demonstrate the advantages of adding the proposed constraint to standard (asymmetric and symmetrized) demons and diffeomorphic demons algorithms through experiments on synthetic images, and real X-ray and 2D/3D brain MRI data. Specifically, the results show that our approach leads to image alignment with more accurate matching of manually defined neuroanatomical structures, better tradeoff between image intensity matching and registration-induced distortion, improved native symmetry, and lower susceptibility to local optima. In summary, the inclusion of this space- and time-varying constraint leads to better image registration along every dimension that we have measured it.

Global image registration using a symmetric block-matching approach

Most medical image registration algorithms suffer from a directionality bias that has been shown to largely impact subsequent analyses. Several approaches have been proposed in the literature to address this bias in the context of nonlinear registration, but little work has been done for global registration. We propose a symmetric approach based on a block-matching technique and least-trimmed square regression. The proposed method is suitable for multimodal registration and is robust to outliers in the input images. The symmetric framework is compared with the original asymmetric block-matching technique and is shown to outperform it in terms of accuracy and robustness. The methodology presented in this article has been made available to the community as part of the NiftyReg open-source package.

Randomized study combining interferon and glatiramer acetate in multiple sclerosis

OBJECTIVE

A double-blind, randomized, controlled study was undertaken to determine whether combined use of interferon β-1a (IFN) 30 μg intramuscularly weekly and glatiramer acetate (GA) 20 mg daily is more efficacious than either agent alone in relapsing-remitting multiple sclerosis.

METHODS

A total of 1,008 participants were randomized and followed until the last participant enrolled completed 3 years. The primary endpoint was reduction in annualized relapse rate utilizing a strict definition of relapse. Secondary outcomes included time to confirmed disability, Multiple Sclerosis Functional Composite (MSFC) score, and magnetic resonance imaging (MRI) metrics.

RESULTS

Combination IFN+GA was not superior to the better of the single agents (GA) in risk of relapse. Both the combination therapy and GA were significantly better than IFN in reducing the risk of relapse. The combination was not better than either agent alone in lessening confirmed Expanded Disability Status Scale progression or change in MSFC over 36 months. The combination was superior to either agent alone in reducing new lesion activity and accumulation of total lesion volumes. In a post hoc analysis, combination therapy resulted in a higher proportion of participants attaining disease activity-free status (DAFS) compared to either single arm, driven by the MRI results.

INTERPRETATION

Combining the 2 most commonly prescribed therapies for multiple sclerosis did not produce a significant clinical benefit over 3 years. An effect was seen on some MRI metrics. In a test of comparative efficacy, GA was superior to IFN in reducing the risk of exacerbation. The extension phase for CombiRx will address whether the observed differences in MRI and DAFS findings predict later clinical differences.

A comprehensive approach to the segmentation of multichannel three-dimensional MR brain images in multiple sclerosis

Accurate classification and quantification of brain tissues is important for monitoring disease progression, measurement of atrophy, and correlating magnetic resonance (MR) measures with clinical disability. Classification of MR brain images in the presence of lesions, such as multiple sclerosis (MS), is particularly challenging. Images obtained with lower resolution often suffer from partial volume averaging leading to false classifications. While partial volume averaging can be reduced by acquiring volumetric images at high resolution, image segmentation and quantification can be technically challenging. In this study, we integrated the brain anatomical knowledge with non-parametric and parametric statistical classifiers for automatically classifying tissues and lesions on high resolution multichannel three-dimensional images acquired on 60 MS brains. The results of automatic lesion segmentation were reviewed by the expert. The agreement between results obtained by the automated analysis and the expert was excellent as assessed by the quantitative metrics, low absolute volume difference percent (36.18 ± 34.90), low average symmetric surface distance (1.64 mm ± 1.30 mm), high true positive rate (84.75 ± 12.69), and low false positive rate (34.10 ± 16.00). The segmented results were also in close agreement with the corrected results as assessed by Bland-Altman and regression analyses. Finally, our lesion segmentation was validated using the MS lesion segmentation grand challenge dataset (MICCAI 2008).

Randomized trial of oral teriflunomide for relapsing multiple sclerosis

BACKGROUND

Teriflunomide is a new oral disease-modifying therapy for relapsing forms of multiple sclerosis.

METHODS

We concluded a randomized trial involving 1088 patients with multiple sclerosis, 18 to 55 years of age, with a score of 0 to 5.5 on the Expanded Disability Status Scale and at least one relapse in the previous year or at least two relapses in the previous 2 years. Patients were randomly assigned (in a 1:1:1 ratio) to placebo, 7 mg of teriflunomide, or 14 mg of teriflunomide once daily for 108 weeks. The primary end point was the annualized relapse rate, and the key secondary end point was confirmed progression of disability for at least 12 weeks.

RESULTS

Teriflunomide reduced the annualized relapse rate (0.54 for placebo vs. 0.37 for teriflunomide at either 7 or 14 mg), with relative risk reductions of 31.2% and 31.5%, respectively (P<0.001 for both comparisons with placebo). The proportion of patients with confirmed disability progression was 27.3% with placebo, 21.7% with teriflunomide at 7 mg (P=0.08), and 20.2% with teriflunomide at 14 mg (P=0.03). Both teriflunomide doses were superior to placebo on a range of end points measured by magnetic resonance imaging (MRI). Diarrhea, nausea, and hair thinning were more common with teriflunomide than with placebo. The incidence of elevated alanine aminotransferase levels (≥1 times the upper limit of the normal range) was higher with teriflunomide at 7 mg and 14 mg (54.0% and 57.3%, respectively) than with placebo (35.9%); the incidence of levels that were at least 3 times the upper limit of the normal range was similar in the lower- and higher-dose teriflunomide groups and the placebo group (6.3%, 6.7%, and 6.7%, respectively). Serious infections were reported in 1.6%, 2.5%, and 2.2% of patients in the three groups, respectively. No deaths occurred.

CONCLUSIONS

Teriflunomide significantly reduced relapse rates, disability progression (at the higher dose), and MRI evidence of disease activity, as compared with placebo. (Funded by Sanofi-Aventis; TEMSO ClinicalTrials.gov number, NCT00134563.).

Automated brain extraction from T2-weighted magnetic resonance images

PURPOSE

To develop and implement an automated and robust technique to extract brain from T2-weighted images.

MATERIALS AND METHODS

Magnetic resonance imaging (MRI) was performed on 75 adult volunteers to acquire dual fast spin echo (FSE) images with fat-saturation technique on a 3T Philips scanner. Histogram-derived thresholds were derived directly from the original images followed by the application of regional labeling, regional connectivity, and mathematical morphological operations to extract brain from axial late-echo FSE (T2-weighted) images. The proposed technique was evaluated subjectively by an expert and quantitatively using Bland-Altman plot and Jaccard and Dice similarity measures.

RESULTS

Excellent agreement between the extracted brain volumes with the proposed technique and manual stripping by an expert was observed based on Bland-Altman plot and also as assessed by high similarity indices (Jaccard: 0.9825 ± 0.0045; Dice: 0.9912 ± 0.0023).

CONCLUSION

Brain extraction using the proposed automated methodology is robust and the results are reproducible.

Estimation of three-dimensional intrinsic dosimetric uncertainties resulting from using deformable image registration for dose mapping

PURPOSE

This article presents a general procedural framework to assess the point-by-point precision in mapped dose associated with the intrinsic uncertainty of a deformable image registration (DIR) for any arbitrary patient.

METHODS

Dose uncertainty is obtained via a three-step process. In the first step, for each voxel in an imaging pair, a cluster of points is obtained by an iterative DIR procedure. In the second step, the dispersion of the points due to the imprecision of the DIR method is used to compute the spatial uncertainty. Two different ways to quantify the spatial uncertainty are presented in this work. Method A consists of a one-dimensional analysis of the modules of the position vectors, whereas method B performs a more detailed 3D analysis of the coordinates of the points. In the third step, the resulting spatial uncertainty estimates are used in combination with the mapped dose distribution to compute the point-by-point dose standard deviation. The process is demonstrated to estimate the dose uncertainty induced by mapping a 62.6 Gy dose delivered on maximum exhale to maximum inhale of a ten-phase four-dimensional lung CT.

RESULTS

For the demonstration lung image pair, the standard deviation of inconsistency vectors is found to be up to 9.2 mm with a mean sigma of 1.3 mm. This uncertainty results in a maximum estimated dose uncertainty of 29.65 Gy if method A is used and 21.81 Gy for method B. The calculated volume with dose uncertainty above 10.00 Gy is 602 cm3 for method A and 1422 cm3 for method B.

CONCLUSIONS

This procedure represents a useful tool to evaluate the precision of a mapped dose distribution due to the intrinsic DIR uncertainty in a patient. The procedure is flexible, allowing incorporation of alternative intrinsic error models.

A reproducible evaluation of ANTs similarity metric performance in brain image registration

The United States National Institutes of Health (NIH) commit significant support to open-source data and software resources in order to foment reproducibility in the biomedical imaging sciences. Here, we report and evaluate a recent product of this commitment: Advanced Neuroimaging Tools (ANTs), which is approaching its 2.0 release. The ANTs open source software library consists of a suite of state-of-the-art image registration, segmentation and template building tools for quantitative morphometric analysis. In this work, we use ANTs to quantify, for the first time, the impact of similarity metrics on the affine and deformable components of a template-based normalization study. We detail the ANTs implementation of three similarity metrics: squared intensity difference, a new and faster cross-correlation, and voxel-wise mutual information. We then use two-fold cross-validation to compare their performance on openly available, manually labeled, T1-weighted MRI brain image data of 40 subjects (UCLA's LPBA40 dataset). We report evaluation results on cortical and whole brain labels for both the affine and deformable components of the registration. Results indicate that the best ANTs methods are competitive with existing brain extraction results (Jaccard=0.958) and cortical labeling approaches. Mutual information affine mapping combined with cross-correlation diffeomorphic mapping gave the best cortical labeling results (Jaccard=0.669±0.022). Furthermore, our two-fold cross-validation allows us to quantify the similarity of templates derived from different subgroups. Our open code, data and evaluation scripts set performance benchmark parameters for this state-of-the-art toolkit. This is the first study to use a consistent transformation framework to provide a reproducible evaluation of the isolated effect of the similarity metric on optimal template construction and brain labeling.

Effect of cocaine on structural changes in brain: MRI volumetry using tensor-based morphometry

Magnetic resonance imaging (MRI) was performed in cocaine-dependent subjects to determine the structural changes in brain compared to non-drug using controls. Cocaine-dependent subjects and controls were carefully screened to rule out brain pathology of undetermined origin. Magnetic resonance images were analyzed using tensor-based morphometry (TBM) and voxel-based morphometry (VBM) without and with modulation to adjust for volume changes during normalization. For TBM analysis, unbiased atlases were generated using two different inverse consistent and diffeomorphic nonlinear registration techniques. Two different control groups were used for generating unbiased atlases. Independent of the nonlinear registration technique and normal cohorts used for creating the unbiased atlases, our analysis failed to detect any statistically significant effect of cocaine on brain volumes. These results show that cocaine-dependent subjects do not show differences in regional brain volumes compared to non-drug using controls.