基于单目视觉的头部磁共振运动伪影矫正系统研究

蒋晗; 陈慧军; 王春尧; 窦佳琦; 黎睿; 徐子茗

doi:10.3969/j.issn.1674-1633.20240370

PDF(2557 KB)

中国医疗设备 ›› 2025, Vol. 40 ›› Issue (2) : 1-5. DOI: 10.3969/j.issn.1674-1633.20240370

研究论著

基于单目视觉的头部磁共振运动伪影矫正系统研究

蒋晗, 陈慧军, 王春尧, 窦佳琦, 黎睿, 徐子茗

作者信息 +

Research on Head Magnetic Resonance Motion Artifacts Correction System Based on Monocular Vision

JIANG Han, CHEN Huijun, WANG Chunyao, DOU Jiaqi, LI Rui, XU Ziming

Author information +

文章历史 +

摘要

目的设计一种基于单目视觉的头部磁共振成像运动矫正系统，以矫正头部磁共振运动伪影，提高成像质量，提供更加精准可靠的成像信息。方法该系统由磁兼容单目相机、光学定位支架、数据采集通信模块组成。通过光学定位支架将单目相机安装于核磁共振腔体内，实时记录人脸信息。基于人脸重建与运动检测算法，实现对人脸的重建与对头部的6自由度刚性运动检测。通过磁共振与相机间交叉标定，利用检测到的运动信息实现磁共振运动伪影的回顾式矫正。结果该系统人脸重建精度达0.57 mm，旋转、平移运动追踪精度分别为0.42°±0.06°和（0.64±0.12）mm，头部磁共振运动伪影显著减少。结论基于单目视觉可实现高精度的6自由度运动追踪与磁共振运动伪影矫正，提高磁共振成像质量，为医生提供更准确的影像信息。

Abstract

Objective To design a head magnetic resonance imaging motion correction system based on monocular vision to correct head magnetic resonance motion artifacts, so as to improve imaging quality and provide more accurate and reliable imaging information. Methods The system was composed of magnetic compatible monocular camera, optical positioning bracket and data acquisition communication module. The monocular camera was installed in the magnetic resonance imaging cavity by optical positioning bracket to record facial information in real-time. Based on facial reconstruction and motion detection algorithm, facial reconstruction and 6-degreeof-freedom rigid motion detection of the head were realized. Through the cross calibration between magnetic resonance and camera, the retrospective correction of magnetic resonance motion artifacts was realized by using the detected motion information. Results The face reconstruction accuracy of the system was 0.57 mm, the tracking accuracy of rotation and translation movements was 0.42°±0.06° and (0.64±0.12) mm, respectively, and the head magnetic resonance motion artifacts were significantly reduced. Conclusion Based on monocular vision, high precision 6-degree-of-freedom motion tracking and magnetic resonance motion artifact correction can be achieved to improve the quality of magnetic resonance imaging and provide more accurate image information for doctors.

导出引用

蒋晗, 陈慧军, 王春尧, 等. 基于单目视觉的头部磁共振运动伪影矫正系统研究[J]. 中国医疗设备, 2025, 40(2): 1-5 https://doi.org/10.3969/j.issn.1674-1633.20240370

JIANG Han, CHEN Huijun, WANG Chunyao, et al. Research on Head Magnetic Resonance Motion Artifacts Correction System Based on Monocular Vision[J]. China Medical Devices, 2025, 40(2): 1-5 https://doi.org/10.3969/j.issn.1674-1633.20240370

中图分类号： R197.39

参考文献

[1] Alexander DC, Dyrby TB, Nilsson M, et al. Imaging brain microstructure with diffusion MRI: practicality and applications[J]. NMR Biomed, 2019, 32(4): e3841.
[2] 孙海峰, 王锐. 功能磁共振成像仪评估运动疲劳后大脑皮质的激活状态[J]. 中国医疗设备, 2022, 37(5): 95-99.
Sun HF, Wang R. Evaluation of activation state of cerebral cortex after exercise fatigue by functional magnetic resonance imaging[J]. China Med Devices, 2022, 37(5): 95-99.
[3] Slipsager JM, Glimberg SL, Sogaard J, et al. Quantifying the financial savings of motion correction in brain MRI: a model-based estimate of the costs arising from patient head motion and potential savings from implementation of motion correction[J]. J Magn Reson Imaging, 2020, 52(3): 731-738.
[4] 曲源, 路青, 蒋杰. 磁共振成像运动伪影控制方法及技术进展[J]. 中国医疗设备, 2022, 37(11): 164-169.
Qu Y, Lu Q, Jiang J. Correction methods and technology progress of magnetic resonance imaging motion artifacts[J]. China Med Devices, 2022, 37(11): 164-169.
[5] Ehman RL, Felmlee JP. Adaptive technique for high-definition MR imaging of moving structures[J]. Radiology, 1989, 173(1): 255-263.
[6] Fu ZW, Wang Y, Grimm RC, et al. Orbital navigator echoes for motion measurements in magnetic resonance imaging[J]. Magn Reson Med, 1995, 34(5): 746-753.
[7] Welch EB, Manduca A, Grimm RC, et al. Spherical navigator echoes for full 3D rigid body motion measurement in MRI[J]. Magn Reson Med, 2002, 47(1): 32-41.
[8] Pipe JG. Motion correction with propeller MRI: application to head motion and free-breathing cardiac imaging[J]. Magn Reson Med, 1999, 42(5): 963-969.
[9] Feng L, Axel L, Chandarana H, et al. XD-GRASP: golden-angle radial MRI with reconstruction of extra motion-state dimensions using compressed sensing[J]. Magn Reson Med, 2016, 75(2): 775-788.
[10] Delattre BM, Heidemann RM, Crowe LA, et al. Spiral demystified[J]. Magn Reson Imaging, 2010, 28(6): 862-881.
[11] Ooi MB, Krueger S, Thomas WJ, et al. Prospective real-time correction for arbitrary head motion using active markers[J]. Magn Reson Med, 2009, 62(4): 943-954.
[12] Zaitsev M, Dold C, Sakas G, et al. Magnetic resonance imaging of freely moving objects: prospective real-time motion correction using an external optical motion tracking system[J]. Neuroimage, 2006, 31(3): 1038-1050.
[13] Andrews-Shigaki BC, Armstrong BS, Zaitsev M, et al. Prospective motion correction for magnetic resonance spectroscopy using single camera Retro-Grate reflector optical tracking[J]. J Magn Reson Imaging, 2011, 33(2): 498-504.
[14] He X, Sun J, Wang Y, et al. Detector-free structure from motion[A]. 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)[C]. 2024.
[15] Qin L, Van Gelderen P, Derbyshire JA, et al. Prospective head-movement correction for high-resolution MRI using an in-bore optical tracking system[J]. Magn Reson Med, 2009, 62(4): 924-934.
[16] Blanz V, Vetter T. A morphable model for the synthesis of 3D faces[A]. Siggraph 99 Conference Proceedings[C]. 1999: 187-194.
[17] Paysan P, Knothe R, Amberg B, et al. A 3D face model for pose and illumination invariant face recognition[A]. 2009 6th IEEE International Conference on Advanced Video and Signal Based Surveillance (AVSS 2009)[C]. Genoa, Italy, IEEE, 2009: 296-301.
[18] Cao C, Weng Y, Zhou S, et al. FaceWarehouse: a 3D facial expression database for visual computing[J]. IEEE Trans Vis Comput Graph, 2014, 20(3): 413-428.
[19] Zhu X, Liu X, Lei Z, et al. Face alignment in full pose range: a 3D total solution[J]. IEEE Trans Pattern Anal Mach Intell, 2019, 41(1): 78-92.
[20] Sanyal S, Bolkart T, Feng HW, et al. Learning to regress 3D face shape and expression from an image without 3D supervision[A]. 2019 IEEE/CVP Conference on Computer Vision and Pattern Recognition: CVPR 2019[C]. Long Beach, California, USA, Institute of Electrical and Electronics Engineers, 2019: 7755-7764.
[21] Feng Y, Wu F, Shao XH, et al. Joint 3D face reconstruction and dense alignment with position map regression network[A]. Computer Vision-ECCV 2018: 15th European Conference[C]. Munich, Germany, Springer, 2018: 557-574.
[22] Li TY, Bolkart T, Black MJ, et al. Learning a model of facial shape and expression from 4D scans[J]. ACM Trans Graph, 2017, 36(6): 194.1-194.17.