Image Registration Performance Research Articles

A light-weight rectangular decomposition large kernel convolution network for deformable medical image registration

The performance and speed of medical image registration have been greatly boosted by advanced deep-learning based methods. However, most current methods are challenged by large deformations between input images, which necessitate a compromise in computational cost to enhance the model’s receptive field and its ability to model long-range spatial relationships for improving registration performance. In order to enhance the performance of registration for images with large deformations at a lower computational cost, in this paper, we propose a light-weight registration model with the ability to model large receptive fields and long-range spatial relationships, named LL-Net. The core components of LL-Net consist of a Rectangular Decomposition Large Kernel Attention (RD-LKA) layer and a Spatial and Channel Fusion Attention (SC-Fusion) layer. The RD-LKA layer utilizes anisotropic depth-wise large kernel convolutions to capture large receptive fields with an extremely low parameter count while modeling long-range spatial relationships. Moreover, the SC-Fusion layer enhances the model’s feature fusion capability and strengthens feature representations at critical locations. Our LL-Net exhibits state-of-the-art performance across multiple datasets. Specifically, it achieves a Dice score of 76.7% and an HD95 of 2.983 mm on the IXI dataset, and a Dice score of 87.8% and an HD95 of 1.042 mm on the OASIS dataset. Experimental results substantiate the efficacy of LL-Net in capturing large receptive fields and modeling long-range spatial relationships. The code for LL-Net is available at https://github.com/BoyOfChu/LL_Net.

Development of Multi-Strip Image Mosaicking for KOMPSAT-3A Images

Abstract. High-resolution satellite imagery has a limitation in terms of coverage area. This limitation presents challenges for extensive-scale analysis at regional or national levels. To maximize the utility of high-resolution satellite imagery, the implementation of image mosaicking techniques is essential. In this paper, we have developed seamline extraction techniques and relative geometric correction optimized for high-resolution satellite imagery. Ultimately, we proposed a multi-strip image mosaicking method for KOMPSAT-3A (Korea Multi-Purpose Satellite-3A) images. We applied the Dijkstra's shortest path algorithm to efficiently extract seamlines. we also performed image registration based on feature matching and homography transformation to correct the relative geometric errors between input images. We conducted experiments with our methods using 29 scenes from KOMPSAT-3A L1G data. The results indicated high relative geometric accuracy, with an average error of 1.63 pixels. Furthermore, we were able to obtain high-quality seamless mosaic images. Our proposed method is expected to enhance the utility of KOMPSAT-3A imagery for large-scale environmental and urban analysis and to provide more accurate and comprehensive data.

On-the-Fly SfM: What you capture is What you get

Abstract. Over the last decades, ample achievements have been made on Structure from Motion (SfM). However, the vast majority of them basically work in an offline manner, i.e., images are firstly captured and then fed together into a SfM pipeline for obtaining poses and sparse point cloud. In this work, on the contrary, we present an on-the-fly SfM: running online SfM while image capturing, the newly taken On-the-Fly image is online estimated with the corresponding pose and points, i.e., what you capture is what you get. Specifically, our approach firstly employs a vocabulary tree that is unsupervised trained using learning-based global features for fast image retrieval of newly fly-in image. Then, a robust feature matching mechanism with least squares (LSM) is presented to improve image registration performance. Finally, via investigating the influence of newly fly-in image’s connected neighboring images, an efficient hierarchical weighted local bundle adjustment (BA) is used for optimization. Extensive experimental results demonstrate that on-the-fly SfM can meet the goal of robustly registering the images while capturing in an online way.

Isocentric fixed angle irradiation-based DRR: a novel approach to enhance x-ray and CT image registration.

Objective.Digitally reconstructed radiography (DRR) plays an important role in the registration of intraoperative x-ray and preoperative CT images. However, existing DRR algorithms often neglect the critical isocentric fixed angle irradiation (IFAI) principle in C-arm imaging, resulting in inaccurate simulation of x-ray images. This limitation degrades registration algorithms relying on DRR image libraries or employing DRR images (DRRs) to train neural network models. To address this issue, we propose a novel IFAI-based DRR method that accurately captures the true projection transformation during x-ray imaging of the human body.Approach.By strictly adhering to the IFAI principle and utilizing known parameters from intraoperative x-ray images paired with CT scans, our method successfully simulates the real projection transformation and generates DRRs that closely resemble actual x-ray images.Main result.Experimental results validate the effectiveness of our IFAI-based DRR method by successfully registering intraoperative x-ray images with preoperative CT images from multiple patients who underwent thoracic endovascular aortic procedures.Significance. The proposed IFAI-based DRR method enhances the quality of DRR images, significantly accelerates the construction of DRR image libraries, and thereby improves the performance of x-ray and CT image registration. Additionally, the method has the generality of registering CT and x-ray images generated by large C-arm devices.

Evaluation and mitigation of deformable image registration uncertainties for MRI-guided adaptive radiotherapy.

We evaluate the performance of a deformable image registration (DIR) software package in registering abdominal magnetic resonance images (MRIs) and then develop a mechanical modeling method to mitigate detected DIR uncertainties. Three evaluation metrics, namely mean displacement to agreement (MDA), DICE similarity coefficient (DSC), and standard deviation of Jacobian determinants (STD-JD), are used to assess the multi-modality (MM), contour-consistency (CC), and image-intensity (II)-based DIR algorithms in the MIM software package, as well as an in-house developed, contour matching-based finite element method (CM-FEM). Furthermore, we develop a hybrid FEM registration technique to modify the displacement vector field of each MIM registration. The MIM and FEM registrations were evaluated on MRIs obtained from 10 abdominal cancer patients. One-tailed Wilcoxon-Mann-Whitney (WMW) tests were conducted to compare the MIM registrations with their FEM modifications. For the registrations performed with the MIM-CC, MIM-MM, MIM-II, and CM-FEM algorithms, their average MDAs are 0.62±0.27, 2.39±1.30, 3.07±2.42, 1.04±0.72mm, and average DSCs are 0.94±0.03, 0.80±0.12, 0.77±0.15, 0.90±0.11, respectively. The p-values of the WMW tests between the MIM registrations and their FEM modifications are less than 0.0084 for STD-JDs and greater than 0.87 for MDA and DSC. Among the three MIM DIR algorithms, MIM-CC shows the smallest errors in terms of MDA and DSC but exhibits significant Jacobian uncertainties in the interior regions of abdominal organs. The hybrid FEM technique effectively mitigates the Jacobian uncertainties in these regions.

A method of UAV visible light remote sensing image registration based on eigenvector technique

As the advancement of unmanned aerial vehicle technology and deep learning technology, many deep learning technologies have been applied in the field of unmanned aerial vehicle image research. To improve the registration precision and efficiency of unmanned aerial vehicle visible light remote sensing images, this study uses the idea of eigenvectors to build an unmanned aerial vehicle visible light remote sensing image registration model. Firstly, the traditional registration model is optimized by multi-branch attention mechanism, and then the final remote sensing image registration model was constructed by combining convolutional neural network and mask prediction module. Testing the performance of the model, and the experiment findings expresses that the registration error of the final model is about 0.05, the registration precision is more than 0.95, and the registration time is within 2s, so it has a good performance of image registration. In conclusion, the image registration model designed by this research institute can improve the existing problems in the current registration field and provide new ideas for improving the image registration effect.

Exploring the performance of implicit neural representations for brain image registration

Pairwise image registration is a necessary prerequisite for brain image comparison and data integration in neuroscience and radiology. In this work, we explore the efficacy of implicit neural representations (INRs) in improving the performance of brain image registration in magnetic resonance imaging. In this setting, INRs serve as a continuous and coordinate based approximation of the deformation field obtained through a multi-layer perceptron. Previous research has demonstrated that sinusoidal representation networks (SIRENs) surpass ReLU models in performance. In this study, we first broaden the range of activation functions to further investigate the registration performance of implicit networks equipped with activation functions that exhibit diverse oscillatory properties. Specifically, in addition to the SIRENs and ReLU, we evaluate activation functions based on snake, sine+, chirp and Morlet wavelet functions. Second, we conduct experiments to relate the hyper-parameters of the models to registration performance. Third, we propose and assess various techniques, including cycle consistency loss, ensembles and cascades of implicit networks, as well as a combined image fusion and registration objective, to enhance the performance of implicit registration networks beyond the standard approach. The investigated implicit methods are compared to the VoxelMorph convolutional neural network and to the symmetric image normalization (SyN) registration algorithm from the Advanced Normalization Tools (ANTs). Our findings not only highlight the remarkable capabilities of implicit networks in addressing pairwise image registration challenges, but also showcase their potential as a powerful and versatile off-the-shelf tool in the fields of neuroscience and radiology.

Prostate volume analysis in image registration for prostate cancer care: a verification study

Combined magnetic resonance imaging (MRI) and positron emission tomography/computed tomography (PET/CT) may enhance diagnosis, aid surgical planning and intra-operative orientation for prostate biopsy and radical prostatectomy. Although PET-MRI may provide these benefits, PET-MRI machines are not widely available. Image fusion of Prostate specific membrane antigen PET/CT and MRI acquired separately may be a suitable clinical alternative. This study compares CT-MR registration algorithms for urological prostate cancer care. Paired whole-pelvis MR and CT scan data were used (n = 20). A manual prostate CTV contour was performed independently on each patients MR and CT image. A semi-automated rigid-, automated rigid- and automated non-rigid registration technique was applied to align the MR and CT data. Dice Similarity Index (DSI), 95% Hausdorff distance (95%HD) and average surface distance (ASD) measures were used to assess the closeness of the manual and registered contours. The automated non-rigid approach had a significantly improved performance compared to the automated rigid- and semi-automated rigid-registration, having better average scores and decreased spread for the DSI, 95%HD and ASD (all p < 0.001). Additionally, the automated rigid approach had similar significantly improved performance compared to the semi-automated rigid registration across all accuracy metrics observed (all p < 0.001). Overall, all registration techniques studied here demonstrated sufficient accuracy for exploring their clinical use. While the fully automated non-rigid registration algorithm in the present study provided the most accurate registration, the semi-automated rigid registration is a quick, feasible, and accessible method to perform image registration for prostate cancer care by urologists and radiation oncologists now.

WarpPINN: Cine-MR image registration with physics-informed neural networks.

The diagnosis of heart failure usually includes a global functional assessment, such as ejection fraction measured by magnetic resonance imaging. However, these metrics have low discriminate power to distinguish different cardiomyopathies, which may not affect the global function of the heart. Quantifying local deformations in the form of cardiac strain can provide helpful information, but it remains a challenge. In this work, we introduce WarpPINN, a physics-informed neural network to perform image registration to obtain local metrics of heart deformation. We apply this method to cine magnetic resonance images to estimate the motion during the cardiac cycle. We inform our neural network of the near-incompressibility of cardiac tissue by penalizing the Jacobian of the deformation field. The loss function has two components: an intensity-based similarity term between the reference and the warped template images, and a regularizer that represents the hyperelastic behavior of the tissue. The architecture of the neural network allows us to easily compute the strain via automatic differentiation to assess cardiac activity. We use Fourier feature mappings to overcome the spectral bias of neural networks, allowing us to capture discontinuities in the strain field. The algorithm is tested on synthetic examples and on a cine SSFP MRI benchmark of 15 healthy volunteers, where it is trained to learn the deformation mapping of each case. We outperform current methodologies in landmark tracking and provide physiological strain estimations in the radial and circumferential directions. WarpPINN provides precise measurements of local cardiac deformations that can be used for a better diagnosis of heart failure and can be used for general image registration tasks. Source code is available at https://github.com/fsahli/WarpPINN.

MD-SGT: Multi-dilation spherical graph transformer for unsupervised medical image registration.

Deformable medical image registration is an essential preprocess step for several clinical applications. Even though the existing convolutional neural network and transformer based methods achieved the promising results, the limited long-range spatial dependence and non-uniform attention span of these models prohibit further improving the registration performance. To deal with this issue, we proposed a multi-dilation spherical graph transformer (MD-SGT), in which the encoder combined the advantages of convolutional and graph transformer blocks to distinguish effectively the differences between the reference and the template images at various scales. Specifically, the features of each voxel were obtained by aggregating the information from its neighbors sampled from different spherical regions with different dilation rates. The implicit convolution inductive bias and long-range uniform attention span induced by such information aggregation manner made the features more representative for registration. Through the qualitative and quantitative comparisons with state-of-the-art methods on two datasets, we demonstrated that combining long-range uniform attention span and inductive bias are beneficial for promoting the image registration performance, with the Dice score, ASD and HD95 being improved at least by 0.5%, 2.2% and 1.1%, respectively.

4D-CT deformable image registration using unsupervised recursive cascaded full-resolution residual networks.

A novel recursive cascaded full-resolution residual network (RCFRR-Net) for abdominal four-dimensional computed tomography (4D-CT) image registration was proposed. The entire network was end-to-end and trained in the unsupervised approach, which meant that the deformation vector field, which presented the ground truth, was not needed during training. The network was designed by cascading three full-resolution residual subnetworks with different architectures. The training loss consisted of the image similarity loss and the deformation vector field regularization loss, which were calculated based on the final warped image and the fixed image, allowing all cascades to be trained jointly and perform the progressive registration cooperatively. Extensive network testing was conducted using diverse datasets, including an internal 4D-CT dataset, a public DIRLAB 4D-CT dataset, and a 4D cone-beam CT (4D-CBCT) dataset. Compared with the iteration-based demon method and two deep learning-based methods (VoxelMorph and recursive cascaded network), the RCFRR-Net achieved consistent and significant gains, which demonstrated that the proposed method had superior performance and generalization capability in medical image registration. The proposed RCFRR-Net was a promising tool for various clinical applications.

Temporal focusing multiphoton microscopy with cross-modality multi-stage 3D U-Net for fast and clear bioimaging.

Temporal focusing multiphoton excitation microscopy (TFMPEM) enables fast widefield biotissue imaging with optical sectioning. However, under widefield illumination, the imaging performance is severely degraded by scattering effects, which induce signal crosstalk and a low signal-to-noise ratio in the detection process, particularly when imaging deep layers. Accordingly, the present study proposes a cross-modality learning-based neural network method for performing image registration and restoration. In the proposed method, the point-scanning multiphoton excitation microscopy images are registered to the TFMPEM images by an unsupervised U-Net model based on a global linear affine transformation process and local VoxelMorph registration network. A multi-stage 3D U-Net model with a cross-stage feature fusion mechanism and self-supervised attention module is then used to infer in-vitro fixed TFMPEM volumetric images. The experimental results obtained for in-vitro drosophila mushroom body (MB) images show that the proposed method improves the structure similarity index measures (SSIMs) of the TFMPEM images acquired with a 10-ms exposure time from 0.38 to 0.93 and 0.80 for shallow- and deep-layer images, respectively. A 3D U-Net model, pretrained on in-vitro images, is further trained using a small in-vivo MB image dataset. The transfer learning network improves the SSIMs of in-vivo drosophila MB images captured with a 1-ms exposure time to 0.97 and 0.94 for shallow and deep layers, respectively.

A discontinuity-preserving regularization for deep learning-based cardiac image registration

Objective. Sliding motion may occur between organs in anatomical regions due to respiratory motion and heart beating. This issue is often neglected in previous studies, resulting in poor image registration performance. A new approach is proposed to handle discontinuity at the boundary and improve registration accuracy. Approach. The proposed discontinuity-preserving regularization (DPR) term can maintain local discontinuities. It leverages the segmentation mask to find organ boundaries and then relaxes the displacement field constraints in these boundary regions. A weakly supervised method using mask dissimilarity loss (MDL) is also proposed. It employs a simple formula to calculate the similarity between the fixed image mask and the deformed moving image mask. These two strategies are added to the loss function during network training to guide the model better to update parameters. Furthermore, during inference time, no segmentation mask information is needed. Main results. Adding the proposed DPR term increases the Dice coefficients by 0.005, 0.009, and 0.081 for three existing registration neural networks CRNet, VoxelMorph, and ViT-V-Net, respectively. It also shows significant improvements in other metrics, including Hausdorff Distance and Average Surface Distance. All quantitative indicator results with MDL have been slightly improved within 1%. After applying these two regularization terms, the generated displacement field is more reasonable at the boundary, and the deformed moving image is closer to the fixed image. Significance. This study demonstrates that the proposed regularization terms can effectively handle discontinuities at the boundaries of organs and improve the accuracy of deep learning-based cardiac image registration methods. Besides, they are generic to be extended to other networks.

Optimized image registration algorithm at subpixel level for Geostationary Operational Environmental Satellite-R Advanced Baseline Imager navigation and registration performance assessment

The image navigation and registration (INR) performance assessment for the Geostationary Operational Environmental Satellite-R (GOES-R) Advanced Baseline Imager (ABI) is an image registration process at the subpixel level, which involves the registration of images from GOES-R ABI with corresponding images from Landsat at the same scene. We present an optimized image registration algorithm for GOES-R ABI INR performance assessment (GAIPA): a gradient descent algorithm with a Sobel edge-enhancement procedure and a normalized lookup table (NLUT) image transformation (GDSN), which is an area-based image registration approach with Pearson correlation as the similarity metric. Sobel edge enhancements increase the sensitivity of Pearson correlations that lead to much improved registration accuracy. NLUT transformations minimize radiance differences between GOES-R ABI images acquired at different times and Landsat reference images on image registration outcomes, which lead to improvements in correlation values and quality of image registration outputs. Evaluations of registration accuracy and quality metrics in GDSN show considerable improvements in computational efficiency, registration accuracy, and quality over the approach implemented on the image INR performance assessment tool set. Pearson correlations can be effective and robust similarity metrics in image registration at the subpixel level with implementing Sobel edge enhancement and NLUT transformation. The statistical analysis of the image registration outputs for GAIPA shows the new algorithm provides considerable improvements in reducing the uncertainties in the registration outputs.

Deep learning based registration for head motion correction in positron emission tomography as a strategy for improved image quantification

Objectives: Positron emission tomography (PET) is affected by various kinds of patient movement during a scan. Frame-by-frame image registration is one of the most practiced motion correction techniques. In recent years, deep learning has shown a remarkable ability to quickly and accurately register images once trained. This paper studies the feasibility of using a deep learning framework to correct 3D positron emission tomography image volumes for head motion in routine positron emission tomography imaging to improve quantification in motion impacted data.Materials and Methods: A neural network was trained with 3D positron emission tomography image volumes in an unsupervised manner to predict transformation parameters required to perform image registration. A multi-step convolutional neural network (CNN) was combined with a spatial transform layer. Pairs of target and source images were used as input to the network. To prepare the training dataset, a previously published TOF-PEPT algorithm was applied to automatically detect static frames where the patient remained in a relatively steady position and transitional frames where they underwent abrupt motion. A single image volume was reconstructed for each static frame. The image reconstructed from the first static frame served as the target image with images from subsequent static frames being used as source images. The trained neural network predicted transformation parameters that could be used to perform frame-by-frame image-based motion correction but also enabled raw listmode positron emission tomography data correction where individual line-of-responses were repositioned. Line profiles and ROIs were drawn across the reconstructed image volumes to compare performance and quantitative results between standard registration tools and the deep learning technique. Corrected volumes were further compared to motion free images quantitatively using Dice indices.Results: In total, one hundred 3D positron emission tomography image volumes were used to train the network. Cross-validation was carried out using a 4:1 ratio for the training and test data. A conventional algorithm for affine registration from the Advanced Normalization Tools (ANTs) software package served as a baseline. To evaluate the correction performance, the mean Dice index and standardized uptake value (SUV) were used. Application of the algorithm to clinical data showed good performance with respect to registration accuracy as well as processing time. The neural network yielded a mean Dice index of ∼0.87 which was similar to the advanced Normalization Tools algorithm and did so ∼3x faster using a multi-core CPU and ∼20x faster with a GPU. Standardized uptake value analysis showed that quantitative results were 30%–60% higher in the motion-corrected images, and the neural network performed better than or close to the advanced Normalization Tools.Conclusion: The aim of this work was to study the quantitative impact of using a data-driven deep learning motion correction technique for positron emission tomography data and assess its performance. The results showed the technique is capable of producing high quality registrations that compensate for patient motion that occurs during a scan and improve quantitative accuracy.

Rotational Distortion and Compensation in Optical Coherence Tomography with Anisotropic Pixel Resolution.

Accurate image registration is essential for eye movement compensation in optical coherence tomography (OCT) and OCT angiography (OCTA). The spatial resolution of an OCT instrument is typically anisotropic, i.e., has different resolutions in the lateral and axial dimensions. When OCT images have anisotropic pixel resolution, residual distortion (RD) and false translation (FT) are always observed after image registration for rotational movement. In this study, RD and FT were quantitively analyzed over different degrees of rotational movement and various lateral and axial pixel resolution ratio (RL/RA) values. The RD and FT provide the evaluation criteria for image registration. The theoretical analysis confirmed that the RD and FT increase significantly with the rotation degree and RL/RA. An image resizing assisting registration (RAR) strategy was proposed for accurate image registration. The performance of direct registration (DR) and RAR for retinal OCT and OCTA images were quantitatively compared. Experimental results confirmed that unnormalized RL/RA causes RD and FT; RAR can effectively improve the performance of OCT and OCTA image registration and distortion compensation.

Learn2Reg: comprehensive multi-task medical image registration challenge, dataset and evaluation in the era of deep learning.

Image registration is a fundamental medical image analysis task, and a wide variety of approaches have been proposed. However, only a few studies have comprehensively compared medical image registration approaches on a wide range of clinically relevant tasks. This limits the development of registration methods, the adoption of research advances into practice, and a fair benchmark across competing approaches. The Learn2Reg challenge addresses these limitations by providing a multi-task medical image registration data set for comprehensive characterisation of deformable registration algorithms. A continuous evaluation will be possible at https:// learn2reg.grand-challenge.org. Learn2Reg covers a wide range of anatomies (brain, abdomen, and thorax), modalities (ultrasound, CT, MR), availability of annotations, as well as intra- and inter-patient registration evaluation. We established an easily accessible framework for training and validation of 3D registration methods, which enabled the compilation of results of over 65 individual method submissions from more than 20 unique teams. We used a complementary set of metrics, including robustness, accuracy, plausibility, and runtime, enabling unique insight into the current state-of-the-art of medical image registration. This paper describes datasets, tasks, evaluation methods and results of the challenge, as well as results of further analysis of transferability to new datasets, the importance of label supervision, and resulting bias. While no single approach worked best across all tasks, many methodological aspects could be identified that push the performance of medical image registration to new state-of-the-art performance. Furthermore, we demystified the common belief that conventional registration methods have to be much slower than deep-learning-based methods.

Incorporating the synthetic CT image for improving the performance of deformable image registration between planning CT and cone-beam CT.

To develop a contrast learning-based generative (CLG) model for the generation of high-quality synthetic computed tomography (sCT) from low-quality cone-beam CT (CBCT). The CLG model improves the performance of deformable image registration (DIR). This study included 100 post-breast-conserving patients with the pCT images, CBCT images, and the target contours, which the physicians delineated. The CT images were generated from CBCT images via the proposed CLG model. We used the Sct images as the fixed images instead of the CBCT images to achieve the multi-modality image registration accurately. The deformation vector field is applied to propagate the target contour from the pCT to CBCT to realize the automatic target segmentation on CBCT images. We calculate the Dice similarity coefficient (DSC), 95 % Hausdorff distance (HD95), and average surface distance (ASD) between the prediction and reference segmentation to evaluate the proposed method. The DSC, HD95, and ASD of the target contours with the proposed method were 0.87 ± 0.04, 4.55 ± 2.18, and 1.41 ± 0.56, respectively. Compared with the traditional method without the synthetic CT assisted (0.86 ± 0.05, 5.17 ± 2.60, and 1.55 ± 0.72), the proposed method was outperformed, especially in the soft tissue target, such as the tumor bed region. The CLG model proposed in this study can create the high-quality sCT from low-quality CBCT and improve the performance of DIR between the CBCT and the pCT. The target segmentation accuracy is better than using the traditional DIR.

A review of deep learning-based deformable medical image registration.

The alignment of images through deformable image registration is vital to clinical applications (e.g., atlas creation, image fusion, and tumor targeting in image-guided navigation systems) and is still a challenging problem. Recent progress in the field of deep learning has significantly advanced the performance of medical image registration. In this review, we present a comprehensive survey on deep learning-based deformable medical image registration methods. These methods are classified into five categories: Deep Iterative Methods, Supervised Methods, Unsupervised Methods, Weakly Supervised Methods, and Latest Methods. A detailed review of each category is provided with discussions about contributions, tasks, and inadequacies. We also provide statistical analysis for the selected papers from the point of view of image modality, the region of interest (ROI), evaluation metrics, and method categories. In addition, we summarize 33 publicly available datasets that are used for benchmarking the registration algorithms. Finally, the remaining challenges, future directions, and potential trends are discussed in our review.

Symmetric transformer-based network for unsupervised image registration

Medical image registration is a fundamental and critical task in medical image analysis. With the rapid development of deep learning, convolutional neural networks (CNNs) have dominated the medical image registration field. Due to the disadvantage of the local receptive field of CNNs, some recent registration methods have focused on using transformers for nonlocal registration. However, the standard transformer has a vast number of parameters and high computational complexity, which means that it can only be applied at the bottom of registration models. As a result, only coarse information is available at the lowest resolution, limiting the contribution of the transformer in these models. To address these challenges, we propose a convolution-based efficient multihead self-attention (CEMSA) block, which reduces the number of parameters of the traditional transformer and captures local spatial context information to reduce semantic ambiguity in the attention mechanism. Based on the proposed CEMSA, we present a novel symmetric transformer-based model (SymTrans). SymTrans employs the transformer blocks in the encoder and the decoder to model the long-range spatial cross-image relevance. We apply SymTrans to the displacement field and diffeomorphic registration. Experimental results show that our proposed method achieves state-of-the-art performance in image registration. Our code is publicly available at https://github.com/MingR-Ma/SymTrans.