浏览全部资源
扫码关注微信
Published： 16 November 2024 ，
移动端阅览
喻莉， 华毅能. 2024. 端到端对称感知对比学习脑室分割算法. 中国图象图形学报， 29(11):3433-3446
Yu Li， Hua Yineng. 2024. End-to-end symmetry-aware-based contrastive learning cerebral ventricle segmentation algorithm. Journal of Image and Graphics， 29(11):3433-3446
目的
2
脑室是人脑重要结构，在临床实践中，其大小、形状变化与多种慢性和急性神经系统疾病息息相关，对脑室的精确分割能够为脑部相关疾病的诊断提供有价值的辅助信息。随着深度学习在医学图像处理领域的迅速发展，医学图像分割任务取得了重大进展。然而，脑室内出血患者的脑室分割问题仍然有待探索。
方法
2
本文聚焦于脑室内出血患者的脑室分割问题，针对其面临的目标遮挡、边界不清晰等问题，提出针对性的脑室分割算法——基于端到端对比学习对称感知的脑室分割网络。该模型首先基于空间转换网络实现自适应图像校正，获取任意角度下输入图像的脑室对称图像。然后通过对比学习算法并结合加权对称损失函数施加对图像的对称性约束。通过上述方法可实现脑室分割网络的端到端训练，上游网络与下游分割任务协同合作。
结果
2
基于不同分割网络模型的实验结果表明，该方案在脑室内出血患者的脑室分割任务上可取得性能提升，该方案按病例和切片评估的Dice系数指标平均增益分别达到1.09%和1.28%。结合本文算法，最优模型按病例评估的DSC（Dice similarity coefficient）系数和召回率分别达到85.17%和84.03%。
结论
2
本文所提出算法对CT（computed tomography）和MR（magnetic resonance）图像的脑室分割均取得了有效提升，对脑室内出血患者相关医学图像分割提升尤为显著，并且本文方法可移植性强，可适用于多种分割网络。
Objective
2
Cerebral ventricles are one of the most prominent cerebral structures. The size and shape changes of the cerebral ventricle are closely associated with diverse acute and chronic neurological diseases. Accurate ventricle segmentation can help diagnose brain-related diseases by providing valuable auxiliary information. However， manual delineation of cerebral ventricles is a time-consuming task； thus， automatic ventricle segmentation is necessary. Fortunately， with the rapid development of deep learning in the field of medical image processing， automatic medical image segmentation has made considerable progress. However， the ventricle segmentation in patients with intraventricular hemorrhage （IVH） remains unexplored. A few studies focus on the ventricle segmentation of patients with IVH.
Method
2
Cerebral ventricle segmentation can be categorized into healthy/normal and IVH cases. Cerebral ventricles in healthy/normal cases are characterized by their high contrast and clear boundaries. The main challenge lies in the segmentation of small-scale cerebral ventricles in some slices. Notably， in healthy/normal cases， cerebral ventricles are not perfectly symmetric； therefore， penalizing a symmetry constraint would be helpful， especially in dealing with the low-contrast small-scale regions. Cerebral ventricle segmentation in healthy/normal cases is generally less challenging. According to the sizes of IVH， the IVH cases are further classified into small- and large-scale IVH cases. For the small-scale IVH cases， though parts of the cerebral ventricles are completely filled by hemorrhages， only the boundary regions would be affected during segmentation. In these cases， the IVH problem would not significantly degrade the segmentation performance because those regions （i.e.， cerebral ventricles filled by IVH） are of high contrast compared to the background， and segmenting the high-contrast regions is roughly equal to cerebral ventricle segmentation. Large-scale IVH cases are the most challenging problem in cerebral ventricle segmentation. Considering the large hemorrhages， the hemorrhages not only cover parts of the cerebral ventricles but also several background regions. All the regions share similar appearances and contrasts. Classifying these regions as background would produce numerous false negatives， while segmenting them as cerebral ventricles would generate quite a few false positives. Therefore， large-scale IVH would poorly affect cerebral ventricle segmentation performance. Based on the above analysis， this study focuses on the cerebral ventricle segmentation problems of patients with IVH and proposes targeted ventricle segmentation methods for the problems of target occlusion and unclear boundaries. The core idea of the proposed framework is to utilize the symmetry of cerebral ventricles as guidance to alleviate the occlusions formed by IVH. Thus， an end-to-end contrastive learning-based symmetry-aware ventricle segmentation network is proposed in the paper. The model first implements adaptive image correction based on spatial transform networks without additional annotations to obtain the ventricle symmetric images of the input images at any position. A symmetry-aware learning loop is then constructed. The symmetric image pairs are simultaneously inputted into the segmentation network. The ventricles predicted by the network are forced to be symmetric by emphasizing the similarity of the segmentation result pairs. Thus， the occlusions formed by IVH could be alleviated by referring to the healthy ventricles. The ventricles are not completely symmetric； thus， pursuing “hard” symmetry during training is infeasible. Therefore， the contrastive learning algorithm is further combined with the weighted symmetry loss function to impose symmetry constraints on the images. The network can be trained end-to-end， enabling the upstream network to collaborate with the downstream segmentation task.
Result
2
Experimental results based on different segmentation network models demonstrate that the approach proposed in this paper can achieve consistent performance improvements in multiple evaluation metrics in the ventricle segmentation task of patients with IVH. The average increase in patient- and slice-wise dice coefficients based on different baseline models when introducing the proposed method is 1.09% and 1.28%， respectively. When evaluated on the patient level， the optimal model achieved a Dice coefficient and recall of 85.17% and 84.03%， respectively， by incorporating the algorithm proposed in this paper. The qualitative results also reveal the superior performance of the proposed algorithm， which achieves smooth boundaries and complete ventricles with fewer false positives.
Conclusion
2
This paper focuses on cerebral ventricle segmentation， especially with the existence of IVH. Compared to cerebral ventricle segmentation in healthy/normal cases， occlusions formed by IVH would make it challenging for segmentation. Based on the symmetry of cerebral ventricles， a symmetry-aware approach combined with contrastive self-supervised learning is introduced. Therefore， the occlusions are effectively alleviated by referring to the healthy/normal parts of the cerebral ventricles. Experimental results on two different datasets demonstrate a notable advancement in ventricle segmentation of computed tomography（CT） and magnetic resonance（MR） images of healthy/normal and IVH cases. Moreover， the IVH cases demonstrate considerable improvement. More importantly， the proposed approach is independent of specific deep learning architectures and introduces no additional computational complexity. Therefore， the method presented in this paper has strong portability and can be applied to various segmentation networks.
脑室分割深度学习脑室内出血（IVH）对称感知端到端网络
cerebral ventricle segmentationdeep learningintraventricular hemorrhage（IVH）symmetry-awareend-to-end network
Atlason H E， Shao M H， Robertsson V， Sigurdsson S， Gudnason V， Prince J L and Ellingsen L M. 2019. Large-scale parcellation of the ventricular system using convolutional neural networks//Proceedings of SPIE 10953， Medical Imaging 2019： Biomedical Applications in Molecular， Structural， and Functional Imaging. San Diego， USA： SPIE： 109530N ［DOI： 10.1117/12.2514590http://dx.doi.org/10.1117/12.2514590］
Carass A， Shao M H， Li X， Dewey B E， Blitz A M， Roy S， Pham D L， Prince J L and Ellingsen L M. 2017. Whole brain parcellation with pathology： validation on ventriculomegaly patients//3rd International Workshop on Patch-Based Techniques in Medical Imaging. Quebec City， Canada： Springer： 20-28 ［DOI： 10.1007/978-3-319-67434-6_3http://dx.doi.org/10.1007/978-3-319-67434-6_3］
Chaitanya K， Erdil E， Karani N and Konukoglu E. 2020. Contrastive learning of global and local features for medical image segmentation with limited annotations//Proceedings of the 34th International Conference on Neural Information Processing Systems. Vancouver， Canada： Curran Associates Inc.： 1052
Dubost F， de Bruijne M， Nardin M， Dalca A V， Donahue K L， Giese A K， Etherton M R， Wu O， de Groot M， Niessen W， Vernooij M， Rost N S and Schirmer M D. 2020. Multi-atlas image registration of clinical data with automated quality assessment using ventricle segmentation. Medical Image Analysis， 63： #101698 ［DOI： 10.1016/j.media.2020.101698http://dx.doi.org/10.1016/j.media.2020.101698］
Fame R M and Lehtinen M K. 2020. Emergence and developmental roles of the cerebrospinal fluid system. Developmental Cell， 52（3）： 261-275 ［DOI： 10.1016/j.devcel.2020.01.027http://dx.doi.org/10.1016/j.devcel.2020.01.027］
Fan Y， Jiang T Z and Evans D J. 2002. Volumetric segmentation of brain images using parallel genetic algorithms. IEEE Transactions on Medical Imaging， 21（8）： 904-909 ［DOI： 10.1109/TMI.2002.803126http://dx.doi.org/10.1109/TMI.2002.803126］
Ferdian E， Boers A M， Beenen L F， Cornelissen B M， Jansen I G， Treurniet K M， Borst J， Majoie C B and Marquering H A. 2017. Automated ventricular system segmentation in CT images of deformed brains due to ischemic and subarachnoid hemorrhagic stroke//5th International Workshop on Molecular Imaging， Reconstruction and Analysis of Moving Body Organs， and Stroke Imaging and Treatment. Québec City， Canada： Springer： 149-157 ［DOI： 10.1007/978-3-319-67564-0_15http://dx.doi.org/10.1007/978-3-319-67564-0_15］
Ghafoorian M， Teuwen J， Manniesing R， de Leeuw F E， van Ginneken B， Karssemeijer N and Platel B. 2018. Student beats the teacher： deep neural networks for lateral ventricles segmentation in brain MR//Proceedings of SPIE 10574， Medical Imaging 2018： Image Processing. Houston， USA： SPIE： 105742U ［DOI： 10.1117/12.2293569http://dx.doi.org/10.1117/12.2293569］
Gu R， Wang G T， Song T， Huang R， Aertsen M， Deprest J， Ourselin S， Vercauteren T and Zhang S T. 2021. CA-Net： comprehensive attention convolutional neural networks for explainable medical image segmentation. IEEE Transactions on Medical Imaging， 40（2）： 699-711 ［DOI： 10.1109/TMI.2020.3035253http://dx.doi.org/10.1109/TMI.2020.3035253］
Hua Y， Li Z Z， Pan J H and Yang X. 2023. Boundary-preserving multi-scale glomerulus segmentation for full-stained kidney slice. Journal of Image and Graphics， 28（11）： 3575-3589
花勇， 李珍珍， 潘建宏， 杨烜. 2023. 边界信息保持的全染色肾脏切片多粒度分割. 中国图象图形学报， 28（11）： 3575-3589 ［DOI： 10.11834/jig.221025http://dx.doi.org/10.11834/jig.221025］
Hua Y N， Yan Z Q， Kuang Z， Zhang H， Deng X B and Yu L. 2022. Symmetry-aware deep learning for cerebral ventricle segmentation with intra-ventricular hemorrhage. IEEE Journal of Biomedical and Health Informatics， 26（10）： 5165-5176 ［DOI： 10.1109/JBHI.2022.3190494http://dx.doi.org/10.1109/JBHI.2022.3190494］
Huff T J， Ludwig P E， Salazar D and Cramer J A. 2019. Fully automated intracranial ventricle segmentation on CT with 2D regional convolutional neural network to estimate ventricular volume. International Journal of Computer Assisted Radiology and Surgery， 14（11）： 1923-1932 ［DOI： 10.1007/s11548-019-02038-5http://dx.doi.org/10.1007/s11548-019-02038-5］
Ironside N， Chen C J， Ding D L， Mayer S A and Connolly Jr E S. 2019. Perihematomal edema after spontaneous intracerebral hemorrhage. Stroke， 50（6）： 1626-1633 ［DOI： 10.1161/STROKEAHA.119.024965http://dx.doi.org/10.1161/STROKEAHA.119.024965］
Jaderberg M， Simonyan K， Zisserman A and Kavukcuoglu K. 2015. Spatial Transformer networks//Proceedings of the 28th International Conference on Neural Information Processing Systems. Montreal， Canada： MIT Press： 2017-2025
Kahle K T， Kulkarni A V， Limbrick D D and Warf B C. 2016. Hydrocephalus in children. The Lancet， 387（10020）： 788-799 ［DOI： 10.1016/S0140-6736（15）60694-8http://dx.doi.org/10.1016/S0140-6736（15）60694-8］
Khan A， Qashqari S and Al-Ali A A. 2013. Accuracy of non-contrast CT brain interpretation by emergency physicians： a cohort study. Pakistan Journal of Medical Sciences， 29（2）： 549-553 ［DOI： 10.12669/pjms.292.3220http://dx.doi.org/10.12669/pjms.292.3220］
Loizou C P， Kyriacou E C， Seimenis I， Pantziaris M， Petroudi S， Karaolis M and Pattichis C S. 2013. Brain white matter lesion classification in multiple sclerosis subjects for the prognosis of future disability. Intelligent Decision Technologies， 7（1）： 3-10 ［DOI： 10.3233/IDT-120147http://dx.doi.org/10.3233/IDT-120147］
Loizou C P， Murray V， Pattichis M S， Seimenis I， Pantziaris M and Pattichis C S. 2011. Multiscale amplitude-modulation frequency-modulation （AM-FM） texture analysis of multiple sclerosis in brain MRI images. IEEE Transactions on Information Technology in Biomedicine， 15（1）： 119-129 ［DOI： 10.1109/TITB.2010.2091279http://dx.doi.org/10.1109/TITB.2010.2091279］
Loizou C P， Pantziaris M， Seimenis I and Pattichis C S. 2009. Brain MR image normalization in texture analysis of multiple sclerosis//Proceedings of the 9th International Conference on Information Technology and Applications in Biomedicine. Larnaka， Cyprus： IEEE： 1-5 ［DOI： 10.1109/ITAB.2009.5394331http://dx.doi.org/10.1109/ITAB.2009.5394331］
Loizou C P， Petroudi S， Seimenis I， Pantziaris M and Pattichis C S. 2015. Quantitative texture analysis of brain white matter lesions derived from T2-weighted MR images in MS patients with clinically isolated syndrome. Journal of Neuroradiology， 42（2）： 99-114 ［DOI： 10.1016/j.neurad.2014.05.006http://dx.doi.org/10.1016/j.neurad.2014.05.006］
Long J， Shelhamer E and Darrell T. 2015. Fully convolutional networks for semantic segmentation//Proceedings of 2015 IEEE Conference on Computer Vision and Pattern Recognition. Boston， USA： IEEE： 3431-3440 ［DOI： 10.1109/CVPR.2015.7298965http://dx.doi.org/10.1109/CVPR.2015.7298965］
Lu L and Qi W M. 2023. Spine CT image segmentation based on Transformer. Journal of Image and Graphics， 28（11）： 3618-3628
卢玲， 漆为民. 2023. 基于Transformer的脊椎CT图像分割. 中国图象图形学报， 28（11）： 3618-3628 ［DOI： 10.11834/jig.221084http://dx.doi.org/10.11834/jig.221084］
Ono K， Iwamoto Y， Chen Y W and Nonaka M. 2020. Automatic segmentation of infant brain ventricles with hydrocephalus in MRI based on 2.5D U-Net and transfer learning. Journal of Image and Graphics， 8（2）： 42-46 ［DOI： 10.18178/joig.8.2.42-46http://dx.doi.org/10.18178/joig.8.2.42-46］
Qian X H， Lin Y， Zhao Y， Yue X Y， Lu B H and Wang J. 2017. Objective ventricle segmentation in brain CT with ischemic stroke based on anatomical knowledge. BioMed Research International， 2017： 8690892 ［DOI： 10.1155/2017/8690892http://dx.doi.org/10.1155/2017/8690892］
Ronneberger O， Fischer P and Brox T. 2015. U-Net： convolutional networks for biomedical image segmentation//Proceedings of the 18th International Conference on Medical Image Computing and Computer-Assisted Intervention. Munich， Germany： Springer： 234-241 ［DOI： 10.1007/978-3-319-24574-4_28http://dx.doi.org/10.1007/978-3-319-24574-4_28］
Schlemper J， Oktay O， Schaap M， Heinrich M， Kainz B， Glocker B and Rueckert D. 2019. Attention gated networks： learning to leverage salient regions in medical images. Medical Image Analysis， 53： 197-207 ［DOI： 10.1016/j.media.2019.01.012http://dx.doi.org/10.1016/j.media.2019.01.012］
Shoykhet M and Clark R S B. 2011. Structure， function， and development of the nervous system//Fuhrman B P and Zimmerman J J， eds. Pediatric Critical Care. 4th ed. Amsterdam， the Netherlands： Elsevier： 783-804 ［DOI： 10.1016/B978-0-323-07307-3.10057-6http://dx.doi.org/10.1016/B978-0-323-07307-3.10057-6］
Xiang T G， Zhang C Y， Liu D N， Song Y， Huang H and Cai W D. 2020. BiO-Net： learning recurrent bi-directional connections for encoder-decoder architecture//Proceedings of the 23rd International Conference on Medical Image Computing and Computer Assisted Intervention. Lima， Peru： Springer： 74-84 ［DOI： 10.1007/978-3-030-59710-8_8http://dx.doi.org/10.1007/978-3-030-59710-8_8］
Zhang X F， Zhang S， Zhang D H and Liu R. 2023. Group attention-based medical image segmentation model. Journal of Image and Graphics， 28（10）： 3231-3242
张学峰， 张胜， 张冬晖， 刘瑞. 2023. 引入分组注意力的医学图像分割模型. 中国图象图形学报， 28（10）： 3231-3242 ［DOI： 10.11834/jig.220748http://dx.doi.org/10.11834/jig.220748］
相关文章
相关作者
相关机构
京公网安备11010802024621