Image registration and appearance adaptation in non-correspondent image regions for new MS lesions detection.
Ontology highlight
ABSTRACT:
SUBMITTER: Andresen J
PROVIDER: S-EPMC9490269 | biostudies-literature | 2022
REPOSITORIES: biostudies-literature
Publications
Image registration and appearance adaptation in non-correspondent image regions for new MS lesions detection.
Andresen Julia J Uzunova Hristina H Ehrhardt Jan J Kepp Timo T Handels Heinz H
Frontiers in neuroscience 20220907
Manual detection of newly formed lesions in multiple sclerosis is an important but tedious and difficult task. Several approaches for automating the detection of new lesions have recently been proposed, but they tend to either overestimate the actual amount of new lesions or to miss many lesions. In this paper, an image registration convolutional neural network (CNN) that adapts the baseline image to the follow-up image by spatial deformations and simulation of new lesions is proposed. Simultane ...[more]
Similar Datasets
Project description:Automatic Non-rigid Histological Image Registration (ANHIR) challenge was organized to compare the performance of image registration algorithms on several kinds of microscopy histology images in a fair and independent manner. We have assembled 8 datasets, containing 355 images with 18 different stains, resulting in 481 image pairs to be registered. Registration accuracy was evaluated using manually placed landmarks. In total, 256 teams registered for the challenge, 10 submitted the results, and 6 participated in the workshop. Here, we present the results of 7 well-performing methods from the challenge together with 6 well-known existing methods. The best methods used coarse but robust initial alignment, followed by non-rigid registration, used multiresolution, and were carefully tuned for the data at hand. They outperformed off-the-shelf methods, mostly by being more robust. The best methods could successfully register over 98% of all landmarks and their mean landmark registration accuracy (TRE) was 0.44% of the image diagonal. The challenge remains open to submissions and all images are available for download.
| S-EPMC7584382 | biostudies-literature
Project description:In the registration of medical images, nonrigid registration targets, images with large displacement caused by different postures of the human body, and frequent variations in image intensity due to physiological phenomena are substantial problems that make medical images less suitable for intensity-based image registration modes. These problems also greatly increase the difficulty and complexity of feature detection and matching for feature-based image registration modes. This research introduces an automatic image registration algorithm for infrared medical images that offers the following benefits: effective detection of feature points in flat regions (cold patterns) that appear due to changes in the human body's thermal patterns, improved mismatch removal through coherent spatial mapping for improved feature point matching, and large-displacement optical flow for optimal transformation. This method was compared with various classical gold standard image registration methods to evaluate its performance. The models were compared for the three key steps of the registration process-feature detection, feature point matching, and image transformation-and the results are presented visually and quantitatively. The results demonstrate that the proposed method outperforms existing methods in all tasks, including in terms of the features detected, uniformity of feature points, matching accuracy, and control point sparsity, and achieves optimal image transformation. The performance of the proposed method with four common image types was also evaluated, and the results verify that the proposed method has a high degree of stability and can effectively register medical images under a variety of conditions.
| S-EPMC9106756 | biostudies-literature
Project description:We introduce an end-to-end deep-learning framework for 3D medical image registration. In contrast to existing approaches, our framework combines two registration methods: an affine registration and a vector momentum-parameterized stationary velocity field (vSVF) model. Specifically, it consists of three stages. In the first stage, a multi-step affine network predicts affine transform parameters. In the second stage, we use a U-Net-like network to generate a momentum, from which a velocity field can be computed via smoothing. Finally, in the third stage, we employ a self-iterable map-based vSVF component to provide a non-parametric refinement based on the current estimate of the transformation map. Once the model is trained, a registration is completed in one forward pass. To evaluate the performance, we conducted longitudinal and cross-subject experiments on 3D magnetic resonance images (MRI) of the knee of the Osteoarthritis Initiative (OAI) dataset. Results show that our framework achieves comparable performance to state-of-the-art medical image registration approaches, but it is much faster, with a better control of transformation regularity including the ability to produce approximately symmetric transformations, and combining affine as well as non-parametric registration.
| S-EPMC7286599 | biostudies-literature
Project description:Background and purposeMR imaging is the key examination in the follow-up of patients with MS, by identification of new high-signal T2 brain lesions. However, identifying new lesions when scrolling through 2 follow-up MR images can be difficult and time-consuming. Our aim was to compare an automated coregistration-fusion reading approach with the standard approach by identifying new high-signal T2 brain lesions in patients with multiple sclerosis during follow-up MR imaging.Materials and methodsThis prospective monocenter study included 94 patients (mean age, 38.9 years) treated for MS with dimethyl fumarate from January 2014 to August 2016. One senior neuroradiologist and 1 junior radiologist checked for new high-signal T2 brain lesions, independently analyzing blinded image datasets with automated coregistration-fusion or the standard scroll-through approach with a 3-week delay between the 2 readings. A consensus reading with a second senior neuroradiologist served as a criterion standard for analyses. A Poisson regression and logistic and γ regressions were used to compare the 2 methods. Intra- and interobserver agreement was assessed by the κ coefficient.ResultsThere were significantly more new high-signal T2 lesions per patient detected with the coregistration-fusion method (7 versus 4, P < .001). The coregistration-fusion method detected significantly more patients with at least 1 new high-signal T2 lesion (59% versus 46%, P = .02) and was associated with significantly faster overall reading time (86 seconds faster, P < .001) and higher reader confidence (91% versus 40%, P < 1 × 10-4). Inter- and intraobserver agreement was excellent for counting new high-signal T2 lesions.ConclusionsOur study showed that an automated coregistration-fusion method was more sensitive for detecting new high-signal T2 lesions in patients with MS and reducing reading time. This method could help to improve follow-up care.
| S-EPMC7655423 | biostudies-literature
Project description:Image registration is a key technique in medical image analysis to estimate deformations between image pairs. A good deformation model is important for high-quality estimates. However, most existing approaches use ad-hoc deformation models chosen for mathematical convenience rather than to capture observed data variation. Recent deep learning approaches learn deformation models directly from data. However, they provide limited control over the spatial regularity of transformations. Instead of learning the entire registration approach, we learn a spatially-adaptive regularizer within a registration model. This allows controlling the desired level of regularity and preserving structural properties of a registration model. For example, diffeomorphic transformations can be attained. Our approach is a radical departure from existing deep learning approaches to image registration by embedding a deep learning model in an optimization-based registration algorithm to parameterize and data-adapt the registration model itself. Source code is publicly-available at https://github.com/uncbiag/registration.
| S-EPMC7286567 | biostudies-literature
Project description:PurposeDeformable image registration (DIR) can benefit from additional guidance using corresponding landmarks in the images. However, the benefits thereof are largely understudied, especially due to the lack of automatic landmark detection methods for three-dimensional (3D) medical images.ApproachWe present a deep convolutional neural network (DCNN), called DCNN-Match, that learns to predict landmark correspondences in 3D images in a self-supervised manner. We trained DCNN-Match on pairs of computed tomography (CT) scans containing simulated deformations. We explored five variants of DCNN-Match that use different loss functions and assessed their effect on the spatial density of predicted landmarks and the associated matching errors. We also tested DCNN-Match variants in combination with the open-source registration software Elastix to assess the impact of predicted landmarks in providing additional guidance to DIR.ResultsWe tested our approach on lower abdominal CT scans from cervical cancer patients: 121 pairs containing simulated deformations and 11 pairs demonstrating clinical deformations. The results showed significant improvement in DIR performance when landmark correspondences predicted by DCNN-Match were used in the case of simulated ( p = 0 e 0 ) as well as clinical deformations ( p = 0.030 ). We also observed that the spatial density of the automatic landmarks with respect to the underlying deformation affect the extent of improvement in DIR. Finally, DCNN-Match was found to generalize to magnetic resonance imaging scans without requiring retraining, indicating easy applicability to other datasets.ConclusionsDCNN-match learns to predict landmark correspondences in 3D medical images in a self-supervised manner, which can improve DIR performance.
| S-EPMC9957682 | biostudies-literature
Project description:To confirm the fully automated rigid image registration (A-RIR) accuracy in postoperative spine stereotactic body radiation therapy (SBRT), we conducted a multicenter non-inferiority study compared to the human rigid image registration (H-RIR). Twenty-eight metastatic cancer patients who underwent postoperative spine SBRT are enrolled-image registration (IR) of planning computed tomography (CT) and CT-myelogram for delineating the spinal cord. The adopted A-RIR workflow is a contour-focused algorithm performing a rigid registration by maximizing normalized mutual information (NMI) restricted to the data contained within the automatically extracted contour. Three radiation oncologists (ROs) from multicenters were prompted to review two blinded registrations and choose one for clinical use. Indistinguishable cases were allowed to vote equivalent, counted A-RIR side. A-RIR is considered non-inferior to H-RIR if the lower limit of the 95% confidence interval (CI) of A-RIR preferable/equivalent is greater than 0.45. We also evaluated the NMI improvement from the baseline and the translational/rotational errors between A-RIR and H-RIR. The A-RIR preferable/equivalent was selected in 21 patients (0.75, 95% CI: 0.55-0.89), demonstrating non-inferiority to H-RIR. The A-RIR's NMI improvement was greater than H-RIR in 24 patients: the mean value ± SD was 0.225 ± 0.115 in A-RIR and 0.196 ± 0.114 in H-RIR (P < 0.001). The absolute translational error was 0.38 ± 0.31 mm. The rotational error was -0.03 ± 0.20, 0.05 ± 0.19, -0.04 ± 0.20 degrees in axial, coronal, and sagittal planes (range: -0.66-0.52). In conclusion, A-RIR shows non-inferior to H-RIR in CT and CT-myelogram registration for postoperative spine SBRT planning.
| S-EPMC8776699 | biostudies-literature
Project description:Over 200 georeferenced registered rephotographic compilations of the Faroe Islands are provided in this dataset. The position of each compilation is georeferenced and thus locatable on a map. Each compilation consists of a historical and a corresponding contemporary image showing the same scene. With steady object features, these two images of the same geolocation are aligned pixel accurately. In the summer of 2022, all contemporary images were photographed by A. Schaffland, while historical images were retrieved from the National Museum of Denmark collections. Images show Faroese landscape and cultural heritage sites, focusing on relevant areas when the historical images were taken, e.g., Kirkjubøur, Tórshavn, and Saksun. Historic images date from the end of the 19th century to the middle of the 20th century. The historical images were taken by scientists, surveyors, archaeologists, and painters. All historical images are in the public domain, have no known rights, or are shared under a CC license. The contemporary images by A. Schaffland are released under CC BY-NC-SA 4.0. The dataset is organized as a GIS project. Historic images, not already georeferenced, were referenced with street view services. All historical images were added to the GIS database, containing camera position, viewing direction, etc. Each compilation can be displayed as an arrow from the camera position along the view direction on a map. Contemporary images were registered to historical images using a specialized tool. None or only a suboptimal rephotograph could be taken for some historical images. These historical images are still added to the database together with all other original images, providing additional data for improvements in rephotography methods in the upcoming years. The resulting image pairs can be used in image registration, landscape change, urban development, and cultural heritage research. Further, the database can be used for public engagement in heritage and as a benchmark for further rephotography and time-series projects.
| S-EPMC9969268 | biostudies-literature
Project description:An invariant feature matching method is proposed as a spatially invariant feature matching approach. Deformation effects, such as affine and homography, change the local information within the image and can result in ambiguous local information pertaining to image points. New method based on dissimilarity values, which measures the dissimilarity of the features through the path based on Eigenvector properties, is proposed. Evidence shows that existing matching techniques using similarity metrics--such as normalized cross-correlation, squared sum of intensity differences and correlation coefficient--are insufficient for achieving adequate results under different image deformations. Thus, new descriptor's similarity metrics based on normalized Eigenvector correlation and signal directional differences, which are robust under local variation of the image information, are proposed to establish an efficient feature matching technique. The method proposed in this study measures the dissimilarity in the signal frequency along the path between two features. Moreover, these dissimilarity values are accumulated in a 2D dissimilarity space, allowing accurate corresponding features to be extracted based on the cumulative space using a voting strategy. This method can be used in image registration applications, as it overcomes the limitations of the existing approaches. The output results demonstrate that the proposed technique outperforms the other methods when evaluated using a standard dataset, in terms of precision-recall and corner correspondence.
| S-EPMC4795769 | biostudies-literature
Project description:BackgroundImage registration is one of the most important and universally useful computational tasks in biomedical image analysis. A recent article by Xing & Qiu (IEEE Transactions on Pattern Analysis and Machine Intelligence, 33(10):2081-2092, 2011) is based on an inappropriately narrow conceptualization of the image registration problem as the task of making two images look alike, which disregards whether the established spatial correspondence is plausible. The authors propose a new algorithm, Nonparametric Local Smoothing (NLS) for image registration, but use image similarities alone as a measure of registration performance, although these measures do not relate reliably to the realism of the correspondence map.ResultsUsing data obtained from its authors, we show experimentally that the method proposed by Xing & Qiu is not an effective registration algorithm. While it optimizes image similarity, it does not compute accurate, interpretable transformations. Even judged by image similarity alone, the proposed method is consistently outperformed by a simple pixel permutation algorithm, which is known by design not to compute valid registrations.ConclusionsThis study has demonstrated that the NLS algorithm proposed recently for image registration, and published in one of the most respected journals in computer science, is not, in fact, an effective registration method at all. Our results also emphasize the general need to apply registration evaluation criteria that are sensitive to whether correspondences are accurate and mappings between images are physically interpretable. These goals cannot be achieved by simply reporting image similarities.
| S-EPMC3740790 | biostudies-literature