Distance Map Supervised Landmark Localization for MR-TRUS Registration Xinrui Song1 Xuanang Xu1 Sheng Xu2 Baris Turkbey3 Bradford J. Wood2 Thomas

2025-04-26 0 0 427.79KB 4 页 10玖币

侵权投诉

Distance Map Supervised Landmark Localization for

MR-TRUS Registration

Xinrui Song1, Xuanang Xu1, Sheng Xu2, Baris Turkbey3, Bradford J. Wood2, Thomas

Sanford4, and Pingkun Yan1

1Department of Biomedical Engineering, Rensselaer Polytechnic Institute, Troy, NY, USA

2Center for Interventional Oncology, Radiology & Imaging Sciences, National Institutes of

Health, Bethesda, MD, USA

3Molecular Imaging Program, National Cancer Institute, National Institutes of Health,

Bethesda, MD, USA

4Department of Urology, The State University of New York Upstate Medical University,

Syracuse, NY, USA

1. DESCRIPTION OF PURPOSE

Image-guided interventional procedures often require registering multi-modal images to visualize and analyze

complementary information. For example, prostate cancer biopsy beneﬁts from fusing transrectal ultrasound

(TRUS) imaging with magnetic resonance (MR) imaging to optimize targeted biopsy. However, cross-modal

image registration is a challenging task. This is especially true when the appearance of two image modalities

are vastly diﬀerent. Since registration quality is most reliably evaluated with target registration error (TRE),

it is sensible to directly make use of the anatomical landmark targets from images. Moreover, while image

modalities and thus textures diﬀer, anatomical landmarks are the only information shared across the moving

and the ﬁxed images. Sun et al.1proposed to perform pre-alignment for MR-TRUS registration with manually

labeled landmarks on both images. However, such a procedure is still far from automatic due to the requirement

of manual input at inference time. Heinrich et al.2made use of a landmark detection method speciﬁcally designed

for lung computed tomography (CT) registration, which is not generalizable to other tasks.

In this work, we propose to explicitly use the landmarks of prostate to guide the MR-TRUS image registration.

We ﬁrst train a deep neural network to automatically localize a set of meaningful landmarks, and then directly

generate the aﬃne registration matrix from the location of these landmarks. For landmark localization, instead

of directly training a network to predict the landmark coordinates, we propose to regress a full-resolution distance

map of the landmark, which is demonstrated eﬀective in avoiding statistical bias to unsatisfactory performance

and thus improving performance. We then use the predicted landmarks to generate the aﬃne transformation

matrix, which outperforms the clinicians’ manual rigid registration by a signiﬁcant margin in terms of TRE.

2. METHODS

Figure 1gives an overview of the proposed method. We ﬁrst extract the location of selected landmarks from

both TRUS and MR images. Then, the two sets of corresponding landmarks are used to calculate an aﬃne

registration matrix that aligns the two images. The acquisition of landmarks is explained in Section 2.1, and the

step that generates the aﬃne registration from corresponding points is explained in Section 2.2.

2.1 Distance map-guided landmark localization

Anatomically important and stable landmarks are key to registration, as it is to our method. We referred to

previous literature1and consulted clinicians to ﬁnally decide on using four landmarks. Two of the landmarks

are the extreme points located on the prostate boundary: the right-most and left-most extreme points observed

from the axial view. The other two (shown in Figure 2) are the entrance point of the urethra into the prostate

at the neck of the bladder, and the point where the urethra exits the prostate at the prostate apex. As the two

Send correspondence to Pingkun Yan (yanp2@rpi.edu).

arXiv:2210.05738v1 [cs.CV] 11 Oct 2022

文档加载中……请稍候！
如果长时间未打开，您也可以点击刷新试试。

下载文档到电脑，查找使用更方便

10 玖币 0人已下载

立即下载

摘要：

DistanceMapSupervisedLandmarkLocalizationforMR-TRUSRegistrationXinruiSong1,XuanangXu1,ShengXu2,BarisTurkbey3,BradfordJ.Wood2,ThomasSanford4,andPingkunYan11DepartmentofBiomedicalEngineering,RensselaerPolytechnicInstitute,Troy,NY,USA2CenterforInterventionalOncology,Radiology&ImagingSciences,NationalIn...

展开>> 收起<<

Distance Map Supervised Landmark Localization for MR-TRUS Registration Xinrui Song1 Xuanang Xu1 Sheng Xu2 Baris Turkbey3 Bradford J. Wood2 Thomas.pdf

共4页,预览1页

还剩页未读，继续阅读

声明：本站为文档C2C交易模式，即用户上传的文档直接被用户下载，本站只是中间服务平台，本站所有文档下载所得的收益归上传人(含作者)所有。玖贝云文库仅提供信息存储空间，仅对用户上传内容的表现方式做保护处理，对上载内容本身不做任何修改或编辑。若文档所含内容侵犯了您的版权或隐私，请立即通知玖贝云文库，我们立即给予删除！

Distance Map Supervised Landmark Localization for MR-TRUS Registration Xinrui Song1 Xuanang Xu1 Sheng Xu2 Baris Turkbey3 Bradford J. Wood2 Thomas

相关推荐

开通VIP享超值会员特权

作者详情

相关内容

热门标签

举报选择: