Image Registration Research Articles

A coarse-to-fine image registration method based on autocorrelation structural difference information

ABSTRACT The automatic registration of synthetic aperture radar (SAR) and optical images is still a challenging problem due to different imaging mechanisms. This letter proposes a coarse-to-fine image registration method that leverages self-similarity structural difference information. In the coarse registration stage, a Scale-Invariant Feature Transformation-based (SIFT-like) method is employed, complemented by an improved Fast Sample Consensus (IFSC) method to eliminate mismatched point pairs by probabilistic and geometric information. This stage ensures robustness against scale and rotational variations. In the fine registration stage, robust feature points are selected by utilizing phase and edge structural information. A descriptor which based on phase consistency and autocorrelation structural difference (ASDPC) is constructed to capture the structural variations between region blocks, and a fine search is carried out within the neighbourhood of the already matched feature points, so as to find more accurate matched feature points and obtain fine registration. The experimental results demonstrate that the proposed method provides robust and accurate registration for optical-to-SAR images.

Enhancing Multispectral Breast Imaging Quality Through Frame Accumulation and Hybrid GA-CPSO Registration

Multispectral transmission imaging has emerged as a promising technique for imaging breast tissue with high resolution. However, the method encounters challenges such as low grayscale, noisy transmission images with weak signals, primarily due to the strong absorption and scattering of light in breast tissue. A common approach to improve the signal-to-noise ratio (SNR) and overall image quality is frame accumulation. However, factors such as camera jitter and respiratory motion during image acquisition can cause frame misalignment, degrading the quality of the accumulated image. To address these issues, this study proposes a novel image registration method. A hybrid approach combining a genetic algorithm (GA) and a constriction factor-based particle swarm optimization (CPSO), referred to as GA-CPSO, is applied for image registration before frame accumulation. The efficiency of this hybrid method is enhanced by incorporating a squared constriction factor (SCF), which speeds up the registration process and improves convergence towards optimal solutions. The GA identifies potential solutions, which are then refined by CPSO to expedite convergence. This methodology was validated on the sequence of breast frames taken at 600 nm, 620 nm, 670 nm, and 760 nm wavelength of light and proved the enhancement of accuracy by various mathematical assessments. It demonstrated high accuracy (99.93%) and reduced registration time. As a result, the GA-CPSO approach significantly improves the effectiveness of frame accumulation and enhances overall image quality. This study explored the groundwork for precise multispectral transmission image segmentation and classification.

Quantifying the spatial distribution of the accumulated dose uncertainty using the novel delta index.

Inter- and intra-fractional anatomical changes during a radiotherapy treatment can cause differences between the initially planned dose and the delivered dose. The total delivered dose can be accumulated over all fractions by using Deformable Image Registration (DIR). However, there is uncertainty in this process which should be accounted for. The aim of this study is to propose a novel metric estimating the spatial distribution of the accumulated dose uncertainty and to evaluate its performance for multi-fraction online adaptive treatments.

Approach: We postulate a new metric, the delta (δ) index, to estimate the uncertainties associated with the dose accumulation process. This metric is calculated for each voxel and takes into account the spatial uncertainty in DIR and local dose differences. For the spatial uncertainty of the DIR, the Distance Discordance Metric (DDM) was used. The accumulated dose and the δ index were determined for ten lung Stereotactic Body Radiation Therapy (SBRT) patients. The δ index was complemented by a more understandable metric, the δ index passing rate, which is the percentage of points satisfying the passing criteria in a region. 

Main Results: The spatial distribution of the δ index and the δ index passing rates showed that voxels failing the criteria were predominantly in lower-dose regions. The mean percentage of voxels passing the criterion increased from 65% to 78%, for threshold doses of 20% and 90% of the prescription doses, respectively. 

Significance: The δ index was postulated to quantify the spatial distribution of the uncertainties associated with the dose accumulation process. The metric gives an intuitive understanding of the reliability of accumulated dose distributions and derived DVH metrics. The performance of the δ index was evaluated for multi-fraction online adaptive treatments, where a case of sub-optimal image registration was identified by the metric.

An unsupervised medical image registration network for intelligent medical education

An unsupervised medical image registration network for intelligent medical education

Analysis of the Relationship Between Scale Invariant Feature Transform Keypoint Properties and Their Invariance to Geometrical Transformation Applied to Cone-Beam Computed Tomography Images

Image registration is a crucial post-processing technique in biomedical imaging, enabling the alignment and integration of images from various sources to facilitate accurate diagnosis, treatment planning, and longitudinal studies. This paper explores the application of Scale Invariant Feature Transform (SIFT), a robust feature-based method for the alignment of biomedical images. SIFT is particularly advantageous due to its invariance to scale, rotation, and affine transformations, making it well-suited for handling the diverse and complex nature of biomedical images. However, SIFT was not initially developed specifically for medical imaging applications, so it is necessary to adapt the algorithm to those kinds of images. In particular, this work was focused on images obtained with Cone-Beam Computed Tomography (CBCT) technology. Besides fine-tuning SIFT parameters on a case-by-case basis, the novelty of this work consists of finding the optimal SIFT parameters on the basis of the keypoints stability. A statistical analysis throughout a dataset of images obtained with CBCT technology was performed to find the best SIFT parameters setting, in terms of computational cost and result quality, compared to default presets.

Deep learning based super-resolution for CBCT dose reduction in radiotherapy.

Cone-beam computed tomography (CBCT) is a crucial daily imaging modality in image-guided and adaptive radiotherapy. However, the use of ionizing radiation in CBCT imaging increases the risk of secondary cancers, which is particularly concerning for pediatric patients. Deep learning super-resolution has shown promising results in enhancing the resolution of photographic and medical images but has not yet been explored in the context of CBCT dose reduction. To facilitate CBCT imaging dose reduction, we propose using an enhanced super-resolution generative adversarial network (ESRGAN) in both the projection and image domains to restore the image quality of low-dose CBCT. An extensive projection database, containing 2997 CBCT scans from head and neck cancer patients, was used to train two different ESRGAN models to generate super-resolution CBCTs. One model operated in the projection domain, using pairs of simulated low-resolution (low-dose) and original high-resolution (high-dose) projections and yielded CBCTSRpro. The other model operated in the image domain, using pairs of axial slices from reconstructed low-resolution and high-resolution CBCTs (CBCTLR and CBCTHR) and resulted in CBCTSRimg. Super-resolution CBCTs were evaluated in terms of image similarity (MAE, ME, PSNR, and SSIM), noise characteristics, spatial resolution, and registration accuracy, using the original CBCT as a reference. To test the perceptual difference between the original and super-resolution CBCT, we performed a visual Turing test. Visually, both super-resolution approaches in the projection and image domains improved the image quality of low-dose CBCTs. This was confirmed by the visual Turing test, that showed low accuracy, sensitivity, and specificity, indicating almost no perceptual difference between CBCTHR and the super-resolution CBCTs. CBCTSRimg (accuracy: 0.55, sensitivity: 0.59, specificity: 0.50) performed slightly better than CBCTSRpro (accuracy: 0.59, sensitivity: 0.61, specificity: 0.57). Image similarity metrics were affected by varying noise levels and did not reflect the visual improvements, with MAE/ME/PSNR/SSIM values of 110.4HU/2.9HU/40.4dB/0.82 for CBCTLR, 136.6HU/-0.4HU/38.6dB/0.77 for CBCTSRpro, and 128.2HU/1.9HU/39.0dB/0.80 for CBCTSRimg. In terms of spatial resolution, quantified by calculating 10% levels of the task transfer function, both CBCTSRpro and CBCTSRimg outperformed CBCTLR and nearly matched the reference CBCTHR (CBCTLR: 0.66lp/mm, CBCTSRpro: 0.88lp/mm, CBCTSRimg: 0.95lp/mm, CBCTHR: 1.01lp/mm). Noise characteristics of CBCTSRimg and CBCTSRpro were comparable to the reference CBCTHR. Registration parameters showed negligible differences for all CBCTs (CBCTLR, CBCTSRpro, CBCTSRimg), with average absolute differences in registration parameters being below 0.4° for rotations and below 0.06mm for translations (CBCTHR as reference). This study demonstrates that deep learning can be a valuable tool for CBCT dose reduction in CBCT-guided radiotherapy by acquiring low-dose CBCTs and restoring the image quality using deep learning super-resolution. The results suggest that higher quality images can be generated when super-resolution is performed in the image domain compared to the projection domain.

CT-augmented digital tomosynthesis image reconstruction in image-guided bronchoscopy interventions.

Transbronchial needle biopsy is crucial for diagnosing lung cancer, yet its efficacy depends on accurately localizing the target lesion and biopsy needle. Digital tomosynthesis (DTS) is considered a promising imaging modality for guiding bronchoscopy procedures due to its low radiation dose and small footprint relative to cone-beam computed tomography (CBCT). However, the image quality of DTS is still not sufficient for an accurate guidance, mainly due to its limited-angleacquisition. Preoperative computed tomography (CT) scans are often performed prior to bronchoscopy interventions for diagnosis or to plan the procedure. The CT images are of high quality and are characterized by a high spatial resolution compared to intraoperative DTS images. These patient-specific prior CT images and the intraoperative DTS images share a fair amount of anatomical information. The main differences only stem from patient positioning and respiratory motion. When these differences are addressed properly, prior CT images augment intraoperative DTS image reconstruction with strong prior knowledge and potentially enhance DTS imagequality. We propose in this work a prior-aided DTS image reconstruction technique leveraging prior CT images to improve DTS image quality. This technique is based on a recently published deformable CT-to-DTS image registration algorithm which is customized for bronchoscopy interventions. The main idea is to register a prior CT image to an intermediate DTS image reconstructed using the standard iterative algebraic reconstruction technique (ART), then to re-reconstruct the DTS image using ART and the registered prior CT image as a firstestimation. The proposed prior-aided reconstruction method was tested on a physical phantom and six patient bronchoscopy datasets. Real DTS data acquired with a pseudo-linear (PL) scan geometry and simulated DTS data generated according to a spherical ellipse (SE) scan geometry were considered. Results evaluated qualitatively by visual inspection and quantitatively by computing Pearson's correlation (PC) with respect to the reference CBCT images suggest significant improvements in image quality using the prior-aided DTS reconstruction compared to the standard zero-initialized ART reconstruction. PC coefficients of the six patient datasets were on average and using a zero-initialized ART reconstruction with SE data and PL data, respectively, and and using the proposed prior-aided reconstruction with SE data and PL data,respectively. While the initial estimation in iterative reconstruction algorithms is often overlooked, we proved that initial estimation is of critical importance in DTS image reconstruction and we have demonstrated the profound advantages of integrating prior CT images in intraoperative DTS image reconstruction. CT-augmented DTS offers a viable alternative to CBCT in guiding bronchoscopy interventions at a fraction of the radiation dose. Further clinical studies are needed to validate improved diagnosticyield.

Introducing a novel sub‐millimeter lung CT image registration error quantitation tool

AbstractBackgroundLung computed tomography (CT) scan image registration is being used for lung function analysis such as ventilation. Given the high sensitivity of functional analyses to image registration errors, an image registration error scoring tool that can measure submillimeter image registration errors is needed.PurposeTo propose an image registration error scoring tool, termed λ, whose spatial sensitivity can be used to quantify image registration errors in steep image gradient regions under realistic noise conditions.Methodsλ compares two images, termed reference and evaluated. The HU and distance scales of both images are normalized by user‐selected scaling criteria. For each voxel in the reference image, the 4D Euclidian distances between the reference voxel and the nearby evaluated voxels are calculated, and the minimum of these distances is . We tested in simulated individual blood vessels comprised of 1, 3, and 5 mm diameter cylinders in 1 × 1 × 1 mm3 voxel images, which were blurred to simulate CT scanner intrinsic resolution and volume averaging. We placed the simulated vessels in a homogeneous background simulating parenchymal tissue density and injected 20, 40, and 60 HU standard deviation Gaussian noise. We used isotropic Gaussian filters with 0.5, 1.0, and 1.5 mm standard deviation kernels to smooth the simulated images. We assessed using reference‐evaluated vessel shifts of −1.0 to 1.0 mm in 0.05 mm steps via rigid translational and rotational deformations. We examined whether tracked the translation vector via its internal spatial component. We restricted to voxels using the angle, termed , between the vector and the normalized spatial‐distance axes, terming the results the restricted‐, , where was hypothesized to be a proxy for image gradient. We determined whether was coincident with the image gradient by examining if the voxels with tracked the evaluated vessels. We used the 95th percentile of , , to determine spatial sensitivity, which we took as a conservative estimate of registration error, by fitting to a modified absolute‐value function for each tested rigid translation, noise level, smoothing kernel, and vessel radius combination. We demonstrated the use of on a clinical example consisting of a set of 25 deformably registered free‐breathing thoracic CT scans. We visually compared the and results against the HU differences between each clinical image pair.ResultsWe found θ to be coincident with the image gradient. We found that ’s spatial component tracked the vessel shifts. We determined the spatial sensitivity limit of to be < 0.2 mm. The noise level and smoothing kernel influenced sensitivity, worsening with increasing noise, and improving with increasing smoothing. For the clinical images, we observed to qualitatively match the absolute difference of intensity in the image pairs and to restrict itself to high gradient regions or regions of visually apparent errors.Conclusion detected sub‐millimeter positioning errors between simulated vessels in the presence of typical CT noise. The noise magnitude and choice of noise smoothing kernel were inversely related to sensitivity, implying that study‐specific tuning of the pre‐smoothing kernel may be required. The demonstrated ability in geometric tests of to detect subvoxel DIR errors warrants further evaluation and testing.

Simultaneous implementation of unrelated tumour sites on the MR Linac: A review of the commissioning process from a radiographer perspective and lessons learned

IntroductionThis work reports on a systematic approach to select MRI sequences, quantify inter-observer image registration variation and determine patient positioning for the clinical implementation of MR-guided adaptive radiotherapy (MRgRT) in patients with oropharyngeal (H&N) and lung cancer. MethodsA total of 30 participants (N=10 H&N and N=10 lung cancer patients and N=10 healthy participants) were scanned on the Elekta Unity Magnetic Resonance Linear Accelerator (MRL). Participant experience questionnaires were used to determine the most appropriate positioning device for lung treatments and tolerability of H&N immobilization devices within the confined MR Linac environment. Visual guided assessments (VGAs) completed by three observers (one oncologist and two radiographers) were used to determine the most suitable tissue weighting (using vendor-provided 3D T1w and T2w sequences) for online image registration. Offline MRI to CT and MRI to MRI rigid registrations were undertaken by nine radiographers using bony and soft tissue matching. Single-factor ANOVA and paired t-tests were utilized to determine the interobserver variation. ResultsBased on oncologist and patient feedback, lung cancer patients would be treated in a vac-bag with their arms by their sides, while H&N cancer patients would be immobilized using a 5-point fixation device and 5-point personalized thermoplastic shell. There was no clear preference for T1w or T2w images in the H&N cohort. However, observers preferred T2w sequences for tumour and organ at risk (OAR) visualization in the lung images. When a bony match was conducted, single-factor ANOVA tests showed no statistically significant differences between all H&N image registration types (p=0.09). For the soft-tissue registrations, T1w-CT and T1w-T1w registrations showed a statistically significant (p=0.01) reduction in inter-observer variability over T2w-CT registrations. Paired t-tests showed no statistically significant differences for bony or soft tissue matches using T1w or T2w sequences to the planning CT in the lung cohorts (p=0.63 and p=0.52, respectively). ConclusionWe describe the systematic approach to the selection of strategies for imaging, immobilization, and online image registration we used for H&N and lung cancer treatments on the MRL. This has facilitated the selection of the most appropriate adaptive MRgRT strategies for treating these sites at our institution.

Multilevel network for large deformation image registration based on feature consistency and flow normalization.

Deformable image registration is an essential technique of medical image analysis, which plays important roles in several clinical applications. Existing deep learning-based registration methods have already achieved promising performance for the registrations with small deformations, while it is still challenging to deal with the large deformation registration due to the limits of the image intensity-similarity-based objectivefunction. To achieve the image registration with large-scale deformations, we proposed a multilevel network architecture FCNet to gradually refine the registration results based on semantic feature consistency constraint and flow normalization (FN)strategy. At each level of FCNet, the architecture is mainly composed to a FeaExtractor, a FN module, and a spatial transformation module. FeaExtractor consists of three parallel streams which are used to extract the individual features of fixed and moving images, as well as their joint features, respectively. Using these features, the initial deformation field is estimated, which passes through a FN module to refine the deformation field based on the difference map of deformation filed between two adjacent levels. This allows the FCNet to progressively improve the registration performance. Finally, a spatial transformation module is used to get the warped image based on the deformation field. Moreover, in addition to the image intensity-similarity-based objective function, a semantic-feature consistency constraint is also introduced, which can further promote the alignments by imposing the similarity between the fixed and warped image features. To validate the effectiveness of the proposed method, we compared our method with the state-of-the-art methods on three different datasets. In EMPIRE10 dataset, 20, 3, and 7 fixed and moving 3D computer tomography (CT) image pairs were used for training, validation, and testing respectively; in IXI dataset, atlas to individual image registration task was performed, with 3D MR images of 408, 58, and 115 individuals were used for training, validation, and testing respectively; in the in-house dataset, patient to atlas registration task was implemented, with the 3D MR images of 94, 3, and 15 individuals being training, validation, and testing sets,respectively. The qualitative and quantitative comparison results demonstrated that the proposed method is beneficial for handling large deformation image registration problems, with the DSC and ASSD improved by at least 1.0% and 25.9% on EMPIRE10 dataset. The ablation experiments also verified the effectiveness of the proposed feature combination strategy, feature consistency constraint, and FNmodule. Our proposed FCNet enables multiscale registration from coarse to fine, surpassing existing SOTA registration methods and effectively handling long-range spatial relationships.

St-RegSeg: an unsupervised registration-based framework for multimodal magnetic resonance imaging stroke lesion segmentation.

Stroke is one of the leading causes of disability and death worldwide. Ischemic stroke accounts for 75-90% of all stroke incidents. Assessing the size and location of the stroke lesion is crucial for treatment decisions, especially those related to urgent vascular reconstruction surgery. Magnetic resonance imaging (MRI) offers excellent soft tissue contrast and multimodal imaging characteristics, which can reflect changes in the physiological functions of brain soft tissues in patients with ischemic stroke. However, using deep learning (DL) techniques for MRI segmentation of stroke lesions still faces many challenges. On the one hand, single-modal segmentation models cannot effectively integrate multimodal information; on the other hand, there is a semantic content drift between multimodal stroke MRI images, leading to lower accuracy in subsequent multimodal image segmentation. To address these issues, we aimed to propose the stroke unsupervised registration and segmentation (St-RegSeg) framework. The St-RegSeg framework integrates an unsupervised registration model, ConvNXMorph, and a segmentation model, nnUNet-v2, enabling both registration and segmentation of multimodal MRI images. The St-RegSeg framework was evaluated on the ISLES'22 dataset from three centers. The St-RegSeg framework demonstrated significant improvements in performance metrics and computational efficiency. Compared to advanced normalization tools (ANTs) [symmetric normalization (SyN)] + nnUNet-v2, the St-RegSeg framework improved the Dice similarity coefficient (DSC) by 25.31% in the registration phase, reduced mean squared error (MSE) by 17.36%, increased normalized cross-correlation (NCC) by 16.06%, and enhanced mutual information (MI) by 17.09%. Additionally, in the segmentation phase, it increased the DSC by 0.84%, and the overall inference speed was increased by 40.91 times. Compared to the suboptimal TransMorph + nnUNet-v2, the St-RegSeg framework improved the DSC by 3.68% in the registration phase, reduced MSE by 8.91%, increased NCC by 8.49%, enhanced MI by 6.18%, and in the segmentation phase, it raised the DSC by 0.5%, with the overall inference speed increased by 2.13 times. The St-RegSeg framework provides a highly effective solution for the registration and segmentation of multimodal MRI images in ischemic stroke cases. Its performance metrics and computational efficiency significantly outperform existing methods, making it a promising tool for clinical applications. The code is open-sourced and available at: https://github.com/Cooper-Gu/St-RegSeg.

Robust thoracic CT image registration with environmental adaptability using dynamic Welsch's function and hierarchical structure-awareness strategy.

Robust registration of thoracic computed tomography (CT) images is strongly impacted by motion during acquisition, high-density objects, and noise, particularly in lower-dose acquisitions. Despite the enhanced registration speed achieved by popular deep learning (DL) methods, their robustness is often neglected. This study aimed to develop a robust thoracic CT image registration algorithm to address the aforementioned issues. A novel, anatomical structure-aware hierarchical registration. By this method, employing a divide-and-conquer approach, dissimilarity metrics, and regularization terms are selected for different regions based on their distinct image features and motion patterns. These terms are then innovatively reconstructed using the Welsch's function, which allows control over the penalty distribution on the loss values. Subsequently, a novel Welsch parameter update strategy is designed for the task of thoracic CT image registration, enabling dynamic sparsity in registration from coarse to fine levels to accommodate various levels of noise and sliding motion. Moreover, the majorization-minimization (MM) algorithm is used to handle the Welsch terms by constructing surrogate functions based on the current variable values for variable update, thereby reducing the complexity of optimization. Experimental results on publicly available deformable image registration lab four-dimensional CT (DIR-Lab 4DCT) and chronic obstructive pulmonary disease (COPD) datasets with and without noise, showed that our proposed method achieves comparable performance to state-of-the-art methods in noise-free scenarios [1.14 and 1.19 mm compared to 1.14 and 1.35 mm target registration errors (TREs)], while demonstrating superior robustness in the presence of noise (1.78 and 2.38 mm compared to 2.00 and 3.31 mm TREs). Ablation studies also validated the effectiveness of each component in the method. A novel and robust algorithm for thoracic CT image registration has been proposed, which has significant potential for valuable clinical applications, including surgical quantitative imaging.

Tibial contact points cannot be used to determine internal-external axial rotation of the native tibiofemoral joint

BackgroundIn the study of tibiofemoral kinematics of the native knee, internal-external (IE) axial rotation is a motion of interest. Locations of contact by the femur on the tibia (termed tibial contact points) have been used to determine IE rotations but such rotations might not be useful due to large error. Hence, our objective was to determine whether tibial contact points are useful in quantifying IE rotations of the native knee. MethodFluoroscopic images of the native knee were analyzed from 25 subjects who performed a weight-bearing deep knee bend. For each subject, 3D bone + cartilage models were created. Following 3D model-to-2D image registration, anterior-posterior (AP) positions of the lowest points and the tibial contact points were computed for each femoral condyle at 0°, 30°, 60°, and 90° of flexion. IE rotations were the angles between lines connecting points in the medial and lateral tibial compartments at different flexion angles. ResultsBased on the lowest points, the tibia rotated internally on the femur primarily during the first 30° of flexion. In this range, mean internal tibial rotation based on tibial contact points was negligible but internal tibial rotation was significantly greater based on lowest points (0° vs 7°, p = 0.0002). At 90° of flexion, the difference was maintained (1.8° vs 8.3°, p = 0.0007). ConclusionWhile tibial contact points are useful in the study of wear of tibial inserts in total knee arthroplasty (TKA), tibial contact points considerably underestimate internal tibial rotation during flexion in the native knee and should not be used to quantify tibiofemoral kinematics.

SPW-TransUNet: three-dimensional computed tomography-cone beam computed tomography image registration with spatial perpendicular window Transformer.

Current medical image registration methods based on Transformer still encounter challenges, including significant local intensity differences and limited computational efficiency when dealing with three-dimensional (3D) computed tomography (CT) and cone beam CT (CBCT) images. These limitations hinder the precise alignment necessary for effective diagnosis and treatment planning. Therefore, the aim of this study is to develop a novel method that overcomes these challenges by enhancing feature interaction and computational efficiency in 3D medical image registration. This paper introduces a novel method that enhances feature interaction within Transformer by computing attention within resizable spatial perpendicular window (SPW). Additionally, it introduces a self-learning mapping control (SLMC) mechanism, which uses a mini convolutional neural network (CNN) to adaptively transform feature vectors into probability vectors. This approach is integrated into the UNet framework, resulting in the SPW-TransUNet. The effectiveness of the SPW-TransUNet is demonstrated through evaluations on two critical 3D medical imaging tasks: CT-CBCT registration and inter-CT registration. We utilized a range of evaluation metrics including Dice similarity coefficient (DICE), structural similarity index measure (SSIM), target registration error (TRE), and negative Jacobian percentage. The validation process involved comparative analysis against established baseline methods using statistical tests to ensure the robustness and reliability of our results. The proposed method demonstrated outstanding performance in the registration of 124 pairs of CT-CBCT lung images from 20 patients, achieving the lowest TRE of 2.16 mm and a minimal negative Jacobian of 0.126. It also recorded the highest SSIM and Dice coefficient of 86.87% and 88.28%, respectively. For the liver CT task involving 150 patients, the method achieved peak SSIM and DICE scores of 76.92% and 85.77%, respectively. Furthermore, ablation studies confirmed the effectiveness of the designed structural components. The SPW-TransUNet offers significant improvements in feature interaction and computational efficiency for medical image registration, providing an effective reference solution for patient and target localization in image-guided radiation therapy.

Fast and highly accurate registration of textile double-sided images: An innovative solution for the calibration of binocular measuring systems

Fast and highly accurate registration of textile double-sided images: An innovative solution for the calibration of binocular measuring systems

Application of intraoral scanning registration implant robot in dental implant surgery.

This paper aims to investigate the application of intraoral scanning and cone beam computed tomography (CBCT) registration implant robot in dental implant surgery. The data of 40 cases with dental defect of robot-assisted implantation from November 2023 to May 2024 were retrospectively analyzed. Before the operation, the intraoral scan data and CBCT data of the positioning markers were automatically fused with the initial CBCT images, and the registration error was calculated. The average registration error of positioning markers was determined during the operation, and the implantation accuracy was analyzed after the operation. The intraoral scan data and CBCT data of 40 patients with dental defect wearing positioning markers were successfully registered with the initial CBCT image, and the registration errors were (0.157±0.026) mm and (0.154±0.033) mm, respectively. Statistical analysis showed no statistical significance between them. The registration errors of the marker was (0.037 3±0.003 6) mm. A total of 55 implants were performed, and the total deviations of the implant point and the apical point were (0.78±0.41) and (0.89±0.28) mm, respectively. The transverse deviations of the implant point and the apical point were (0.44±0.36) and (0.58±0.25) mm, respectively. The depth deviations of the implant point and the apical point were (0.51±0.32) and (0.54±0.36) mm, respectively. The deviation of the implant angle was 1.24°±0.67°. The fusion technology based on intraoral scanning and CBCT registration can meet the accuracy requirements of preoperative registration of oral implant robots. The technology increases the choice of registration methods before robot-assisted dental implant surgery and reduces the multiple radiation exposuresof the patient.

NCMNet: Neighbor Consistency Mining Network for Two-View Correspondence Pruning.

Correspondence pruning plays a crucial role in a variety of feature matching based tasks, which aims at identifying correct correspondences (inliers) from initial ones. Seeking consistent k-nearest neighbors in both coordinate and feature spaces is a prevalent strategy employed in previous approaches. However, the vicinity of an inlier contains numerous irregular false correspondences (outliers), which leads them to mistakenly become neighbors according to the similarity constraint of nearest neighbors. To tackle this issue, we propose a global-graph space to seek consistent neighbors with similar graph structures. This is achieved by using a global connected graph to explicitly render the affinity relationship between correspondences based on the spatial and feature consistency. Furthermore, to enhance the robustness of method for various matching scenes, we develop a neighbor consistency block to adequately leverage the potential of three types of neighbors. The consistency can be progressively mined by sequentially extracting intra-neighbor context and exploring inter-neighbor interactions. Ultimately, we present a Neighbor Consistency Mining Network (NCMNet) to estimate the parametric models and remove outliers. Extensive experimental results demonstrate that the proposed method outperforms other state-of-the-art methods on various benchmarks for two-view geometry estimation. Meanwhile, four extended tasks, including remote sensing image registration, point cloud registration, 3D reconstruction, and visual localization, are conducted to test the generalization ability.

Evaluation of pelvic MRI-to-CT deformable registration for adaptive MR-guided particle therapy

PurposeWe aim to assess the MRI-to-CT deformable image registration (DIR) quality of our treatment planning system in the pelvic region, as the first step of an online MRI-guided particle therapy clinical workflow. Materials and MethodsUsing two different DIR algorithms, ANACONDA, the DIR algorithm incorporated in RayStation, and Elastix, an open source registration software, we retrospectively assessed the quality of the deformed CT (dCT) generation in the pelvic region for five patients. T1- and T2-weighted daily control MRI acquired prior to treatment delivery were used for the DIR. We compared the contours automatically mapped on the dCT against the manual contours on the MRI (ground-truth), by calculating the Dice similarity coefficients (DSC) and mean distances to agreement (MDA) for OAR, targets and outer contour. We assessed the dosimetric impact of the DIR on the clinical treatment plans, comparing the dose volume histograms (DVHs) and the value of the clinical goals achieved for each dCT. The water equivalent path lengths (WEPL) and dose range 80% (R80%) maps were compared by casting on the beams' eye view. ResultsThe T1 sequences performed better for the DIR with ANACONDA compared against the T2. ANACONDA's performance agreed with Elastix. The bladder and rectum led to the worst agreement. For the remaining structures analysed, DSC above 0.80 were obtained. Maximum median deviations of 7.1 mm and 2.1 mm were observed for WEPL and R80%, respectively, on the PTV. ConclusionThis works shows a good agreement on the DIR quality achieved with ANACONDA for the structures in the beams' path. By comparing the R80% generated with ANACONDA and Elastix, we give a first quantification of the uncertainties to be considered in an online MRI-guided particle therapy workflow for pelvic treatment.

An open-source software for the simulation of fixational eye movements

Abstract Laser-scanning confocal microscopy of the human cornea acquires in vivoimages of corneal tissues with cellular resolution, which offers diagnostic potential for a variety of diseases. However, involuntary fixational eye movements induce motion artefacts that pose a challenge for accurate morphometric analysis and particularly for preceding image data fusion steps. Different image registration algorithms promise to eliminate such artefacts, but objective, quantitative evaluation of the registration results is difficult because of the lack of ground truth data. Simulation approaches offer a possibility to close this gap. Existing open-source software for the simulation of fixational eye movements is currently limited to creating drift movement. The present contribution proposes complementary models for microsaccade and tremor components to provide complete simulated fixational eye movement paths. In addition, it reports results from an extensive examination of different model parameter combinations. It is shown that the microsaccade model is capable of reproducing intermicrosaccade intervals that closely resemble experimental data. A software implementation is provided as open-source Python modules.

Integration of a lightweight and table-mounted robotic alignment tool with automated patient-to-image registration using robotic cone-beam CT for intracranial biopsies and stereotactic electroencephalography.

Robotics in neurosurgery is becoming increasingly prevalent. The integration of intraoperative imaging for patient registration into workflows of newer robotic systems enhances precision and has further driven their widespread adoption. In this study, the authors report on a lightweight, table-mounted robotic system integrating robotic cone-beam CT (CB-CT) for automated patient registration in cranial biopsies and stereotactic electroencephalography (sEEG). This prospective cohort study included patients who underwent stereotactic biopsy or sEEG with the Cirq system from January 2023 to August 2024. For patient-to-image registration, an external registration matrix was secured near the patient's head before conducting CB-CT with robotic Artis Pheno. CT was then fused with preoperative planning MRI and used as the navigation dataset. Demographic and clinical data were evaluated, and entry and target errors, as well as vector deviation of sEEG electrodes, were assessed and compared with those of patients who underwent biopsies and sEEG with the frameless VarioGuide system. In 26 Cirq-assisted surgical procedures, robotic CB-CT was used for image registration in 20 cases. Of these, 15 were biopsies (mean ± SD 7 ± 1 specimens) and 5 were sEEG with 31 depth electrodes, compared to 29 VarioGuide biopsies and 3 VarioGuide sEEG cases with 25 electrodes. The mean age was 56 ± 19 years, with a male/female ratio of 1.9:1. Lesion size averaged 19 ± 17 cm3 on T1-weighted imaging and 61 ± 53 cm3 on T2-weighted imaging for Cirq and 14 ± 14 cm3 and 68 ± 47 cm3 for VarioGuide. The mean surgical times were 117 ± 34 minutes for biopsy and 269 ± 54 minutes for sEEG in the Cirq group, with skin-to-skin times of 40 ± 23 minutes for biopsy and 208 ± 74 minutes for sEEG; in comparison, surgical times of 78 ± 21 minutes for biopsy and 218 ± 33 minutes for sEEG were reported with VarioGuide, with skin-to-skin times of 34 ± 13 and 158 ± 27 minutes. No complications occurred. The mean dosage area product was 983 ± 351 µGym2 for biopsies and 1772 ± 968 µGym2 for sEEG. Cirq-assisted sEEG electrodes had mean entry and target errors of 1.4 ± 1.2 mm and 2.6 ± 1.6 mm, compared to 5.3 ± 3.3 mm and 6.5 ± 2.8 mm with VarioGuide. Mean vector deviation was 1.6 ± 0.9 mm with Cirq versus 4.9 ± 2.9 mm with VarioGuide. The integration of a lightweight, table-mounted robotic alignment tool with intraoperative CB-CT for automated patient-to-image registration enables high precision and a seamless workflow. This combination is safe, has a manageable learning curve, and holds potential to replace traditional frame-based and frameless procedures. Its efficiency and accuracy are likely to contribute to the increasing adoption of robotics in neurosurgery.