Pseudo-Rendering for Resolution and Topology-Invariant Cortical Parcellation.

Pablo Blasco Fernandez, Karthik Gopinath, John Williams-Ramirez, Rogeny Herisse, Lucas J Deden-Binder, Dina Zemlyanker, Theressa Connors, Liana Kozanno, Derek Oakley, Bradley Hyman, Sean I Young, Juan Eugenio Iglesias
{"title":"Pseudo-Rendering for Resolution and Topology-Invariant Cortical Parcellation.","authors":"Pablo Blasco Fernandez, Karthik Gopinath, John Williams-Ramirez, Rogeny Herisse, Lucas J Deden-Binder, Dina Zemlyanker, Theressa Connors, Liana Kozanno, Derek Oakley, Bradley Hyman, Sean I Young, Juan Eugenio Iglesias","doi":"10.1007/978-3-031-73290-4_8","DOIUrl":null,"url":null,"abstract":"<p><p>Parcellation of mesh models for cortical analysis is a central problem in neuroimaging. Most classical and deep learning methods have requisites in terms of mesh topology, requiring inputs that are homeomorphic to a sphere (i.e., no holes or handles). Topology correction algorithms do exist, but their computational complexity is quadratic with the size of the topological defects - sometimes hours, effectively precluding segmentation of incorrect meshes, including those derived from imperfect segmentations or obtained from inherently noisy modalities like surface scanning. Furthermore, deep learning mesh segmentation also struggles surface scans of brains because they are relatively nondescript and require modeling of longer-range dependencies. Here we propose \"pseudo-render-inverse-render\" (PRIR), a novel perspective on cortical mesh parcellation that effectively reframes the problem as a 2D segmentation task using an direct-inverse rendering framework. Our approach: <i>(i)</i> renders the mesh from a number of perspectives, projecting the three components of the face normal vectors to a three-channel image; <i>(ii)</i> segments these images with U-Nets; <i>(iii)</i> maps the 2D segmentations back to vertices (inverse rendering); and <i>(iv)</i> aggregates the information from multiple views, postprocessing the output with a Markov Random Field to ensure smoothness and segmentation of occluded areas. PRIR is not affected by mesh topology and easily captures long-range dependencies with the U-Nets. Our results demonstrate: state-of-the-art accuracy on topologically correct white matter meshes; equally accurate performance on simulated surface scans; and robust segmentation of real surface scans.</p>","PeriodicalId":74092,"journal":{"name":"Machine learning in medical imaging. MLMI (Workshop)","volume":"15242 ","pages":"74-84"},"PeriodicalIF":0.0000,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12398396/pdf/","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Machine learning in medical imaging. MLMI (Workshop)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1007/978-3-031-73290-4_8","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"2024/10/23 0:00:00","PubModel":"Epub","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0

Abstract

Parcellation of mesh models for cortical analysis is a central problem in neuroimaging. Most classical and deep learning methods have requisites in terms of mesh topology, requiring inputs that are homeomorphic to a sphere (i.e., no holes or handles). Topology correction algorithms do exist, but their computational complexity is quadratic with the size of the topological defects - sometimes hours, effectively precluding segmentation of incorrect meshes, including those derived from imperfect segmentations or obtained from inherently noisy modalities like surface scanning. Furthermore, deep learning mesh segmentation also struggles surface scans of brains because they are relatively nondescript and require modeling of longer-range dependencies. Here we propose "pseudo-render-inverse-render" (PRIR), a novel perspective on cortical mesh parcellation that effectively reframes the problem as a 2D segmentation task using an direct-inverse rendering framework. Our approach: (i) renders the mesh from a number of perspectives, projecting the three components of the face normal vectors to a three-channel image; (ii) segments these images with U-Nets; (iii) maps the 2D segmentations back to vertices (inverse rendering); and (iv) aggregates the information from multiple views, postprocessing the output with a Markov Random Field to ensure smoothness and segmentation of occluded areas. PRIR is not affected by mesh topology and easily captures long-range dependencies with the U-Nets. Our results demonstrate: state-of-the-art accuracy on topologically correct white matter meshes; equally accurate performance on simulated surface scans; and robust segmentation of real surface scans.

用于分辨率和拓扑不变皮质分割的伪渲染。
用于皮质分析的网格模型的分割是神经影像学中的一个核心问题。大多数经典和深度学习方法在网格拓扑方面都有要求,要求输入与球体同态(即,没有孔或手柄)。拓扑校正算法确实存在,但它们的计算复杂度与拓扑缺陷的大小是二次的——有时是几个小时,有效地排除了不正确网格的分割,包括那些来自不完美分割或从表面扫描等固有噪声模式中获得的网格。此外,深度学习网格分割也很难对大脑进行表面扫描,因为它们相对来说难以描述,而且需要对更远距离的依赖关系进行建模。在这里,我们提出了“伪渲染-反渲染”(PRIR),这是一种关于皮质网格分割的新视角,它有效地将问题重新定义为使用直接逆渲染框架的2D分割任务。我们的方法:(i)从多个角度渲染网格,将面部法向量的三个组成部分投影到三通道图像;(ii)用U-Nets对这些图像进行分割;(iii)将2D分割映射回顶点(反向渲染);(iv)聚合来自多个视图的信息,用马尔可夫随机场对输出进行后处理,以确保遮挡区域的平滑和分割。PRIR不受网格拓扑结构的影响,并且很容易捕获与U-Nets的远程依赖关系。我们的研究结果表明:在拓扑正确的白质网格上具有最先进的精度;同样准确的模拟表面扫描性能;和真实表面扫描的鲁棒分割。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 求助全文
来源期刊
自引率
0.00%
发文量
0
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:604180095
Book学术官方微信