Abstract:Background : Image registration technology has become an important medical image preprocessing step with the wide application of computer‐aided diagnosis technology in various medical image analysis tasks. Purpose : We propose a multiscale feature fusion registration based on deep learning to achieve the accurate registration and fusion of head magnetic resonance imaging (MRI) and solve the problem that general registration methods cannot handle the complex spatial information and position information of head MRI. Methods : Our proposed multiscale feature fusion registration network consists of three sequentially trained modules. The first is an affine registration module that implements affine transformation; the second is to realize non‐rigid transformation, a deformable registration module composed of top‐down and bottom‐up feature fusion subnetworks in parallel; and the third is a deformable registration module that also realizes non‐rigid transformation and is composed of two feature fusion subnetworks in series. The network decomposes the deformation field of large displacement into multiple deformation fields of small displacement by multiscale registration and registration, which reduces the difficulty of registration. Moreover, multiscale information in head MRI is learned in a targeted manner, which improves the registration accuracy, by connecting the two feature fusion subnetworks. Results : We used 29 3D head MRIs for training and seven volumes for testing and calculated the values of the registration evaluation metrics for the new algorithm to register anterior and posterior lateral pterygoid muscles. The Dice similarity coefficient was 0.745±0.021, the Hausdorff distance was 3.441±0.935 mm, the Average surface distance was 0.738±0.098 mm, and the Standard deviation of the Jacobian matrix was 0.425±0.043. Our new algorithm achieved a higher registration accuracy compared with state‐of‐the‐art registration methods. Conclusions : Our proposed multiscale feature fusion registration network can realize end‐to‐end deformable registration of 3D head MRI, which can effectively cope with the characteristics of large deformation displacement and the rich details of head images and provide reliable technical support for the diagnosis and analysis of head diseases. This article is protected by copyright. All rights reserved

Minimum Spanning Tree Fusing Uniform Sub-Sampling Points and High-Dimensional Features for Medical Image Registration

Minimum Spanning Tree Hierarchically Fusing Multi-feature Points and High-Dimensional Features for Medical Image Registration.

Minimum Spanning Tree Fusing Multi-Salient Points Hierarchically For Multi-Modality Image Registration

Minimum Spanning Tree Integrated with Gradient Information for Medical Image Registration

Medical Image Elastic Registration Based on Contour and Feature Points

Non-rigid Registration of Biomedical Image for Radiotherapy Based on Adaptive Feature Density Flow

Registration of Medical Images Using Points and Curves

Local Registration of Medical Images Using Feature Point and Intensity Information

Non-rigid Registration of Multimodality Medical Images Based on ITK

DoG keypoints medical image registration based on minimum spanning tree

Entropic graph estimation integrated with SIFT features for medical image non-rigid registration

Anatomically Constrained and Attention-Guided Deep Feature Fusion for Joint Segmentation and Deformable Medical Image Registration.

A hybrid registration of medical images using intensity information and landmark points

Complementary scale space keypoint medical image registration based on Rényi entropy

Multi-Modality Medical Image Registration Based on Regional Joint Rényi Entropy

Affine medical image registration with fusion feature mapping in local and global

Multiscale feature fusion network for 3D head MRI image registration

3D Biological/Biomedical Image Registration with enhanced Feature Extraction and Outlier Detection

Transformer-based 2D/3D medical image registration for X-ray to CT via anatomical features

Biological Fiducial Point Based Registration for Multiple Brain Tissues Reconstructed from Different Imaging Modalities

Manifold-based Feature Point Matching for Multi-Modal Image Registration.