Collaborative regression-based anatomical landmark detection

Yaozong Gao, Dinggang Shen

Research output: Contribution to journalArticle

10 Citations (Scopus)

Abstract

Anatomical landmark detection plays an important role in medical image analysis, e.g. for registration, segmentation and quantitative analysis. Among the various existing methods for landmark detection, regression-based methods have recently attracted much attention due to their robustness and efficiency. In these methods, landmarks are localised through voting from all image voxels, which is completely different from the classification-based methods that use voxel-wise classification to detect landmarks. Despite their robustness, the accuracy of regression-based landmark detection methods is often limited due to (1) the inclusion of uninformative image voxels in the voting procedure, and (2) the lack of effective ways to incorporate inter-landmark spatial dependency into the detection step. In this paper, we propose a collaborative landmark detection framework to address these limitations. The concept of collaboration is reflected in two aspects. (1) Multi-resolution collaboration. A multi-resolution strategy is proposed to hierarchically localise landmarks by gradually excluding uninformative votes from faraway voxels. Moreover, for informative voxels near the landmark, a spherical sampling strategy is also designed at the training stage to improve their prediction accuracy. (2) Inter-landmark collaboration. A confidence-based landmark detection strategy is proposed to improve the detection accuracy of 'difficult-to-detect' landmarks by using spatial guidance from 'easy-to-detect' landmarks. To evaluate our method, we conducted experiments extensively on three datasets for detecting prostate landmarks and head & neck landmarks in computed tomography images, and also dental landmarks in cone beam computed tomography images. The results show the effectiveness of our collaborative landmark detection framework in improving landmark detection accuracy, compared to other state-of-the-art methods.

Original languageEnglish
Pages (from-to)9377-9401
Number of pages25
JournalPhysics in Medicine and Biology
Volume60
Issue number24
DOIs
Publication statusPublished - 2015 Nov 18

Fingerprint

Politics
Spatial Analysis
Cone-Beam Computed Tomography
Prostate
Tooth
Neck
Head
Tomography
Datasets

Keywords

  • image analysis
  • image processing
  • image registration
  • landmark detection
  • landmark-based analysis
  • machine learning
  • prostate localization

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging
  • Radiological and Ultrasound Technology

Cite this

Collaborative regression-based anatomical landmark detection. / Gao, Yaozong; Shen, Dinggang.

In: Physics in Medicine and Biology, Vol. 60, No. 24, 18.11.2015, p. 9377-9401.

Research output: Contribution to journalArticle

@article{a1c76936c5144ca7a9ec2053dcb49810,
title = "Collaborative regression-based anatomical landmark detection",
abstract = "Anatomical landmark detection plays an important role in medical image analysis, e.g. for registration, segmentation and quantitative analysis. Among the various existing methods for landmark detection, regression-based methods have recently attracted much attention due to their robustness and efficiency. In these methods, landmarks are localised through voting from all image voxels, which is completely different from the classification-based methods that use voxel-wise classification to detect landmarks. Despite their robustness, the accuracy of regression-based landmark detection methods is often limited due to (1) the inclusion of uninformative image voxels in the voting procedure, and (2) the lack of effective ways to incorporate inter-landmark spatial dependency into the detection step. In this paper, we propose a collaborative landmark detection framework to address these limitations. The concept of collaboration is reflected in two aspects. (1) Multi-resolution collaboration. A multi-resolution strategy is proposed to hierarchically localise landmarks by gradually excluding uninformative votes from faraway voxels. Moreover, for informative voxels near the landmark, a spherical sampling strategy is also designed at the training stage to improve their prediction accuracy. (2) Inter-landmark collaboration. A confidence-based landmark detection strategy is proposed to improve the detection accuracy of 'difficult-to-detect' landmarks by using spatial guidance from 'easy-to-detect' landmarks. To evaluate our method, we conducted experiments extensively on three datasets for detecting prostate landmarks and head & neck landmarks in computed tomography images, and also dental landmarks in cone beam computed tomography images. The results show the effectiveness of our collaborative landmark detection framework in improving landmark detection accuracy, compared to other state-of-the-art methods.",
keywords = "image analysis, image processing, image registration, landmark detection, landmark-based analysis, machine learning, prostate localization",
author = "Yaozong Gao and Dinggang Shen",
year = "2015",
month = "11",
day = "18",
doi = "10.1088/0031-9155/60/24/9377",
language = "English",
volume = "60",
pages = "9377--9401",
journal = "Physics in Medicine and Biology",
issn = "0031-9155",
publisher = "IOP Publishing Ltd.",
number = "24",

}

TY - JOUR

T1 - Collaborative regression-based anatomical landmark detection

AU - Gao, Yaozong

AU - Shen, Dinggang

PY - 2015/11/18

Y1 - 2015/11/18

N2 - Anatomical landmark detection plays an important role in medical image analysis, e.g. for registration, segmentation and quantitative analysis. Among the various existing methods for landmark detection, regression-based methods have recently attracted much attention due to their robustness and efficiency. In these methods, landmarks are localised through voting from all image voxels, which is completely different from the classification-based methods that use voxel-wise classification to detect landmarks. Despite their robustness, the accuracy of regression-based landmark detection methods is often limited due to (1) the inclusion of uninformative image voxels in the voting procedure, and (2) the lack of effective ways to incorporate inter-landmark spatial dependency into the detection step. In this paper, we propose a collaborative landmark detection framework to address these limitations. The concept of collaboration is reflected in two aspects. (1) Multi-resolution collaboration. A multi-resolution strategy is proposed to hierarchically localise landmarks by gradually excluding uninformative votes from faraway voxels. Moreover, for informative voxels near the landmark, a spherical sampling strategy is also designed at the training stage to improve their prediction accuracy. (2) Inter-landmark collaboration. A confidence-based landmark detection strategy is proposed to improve the detection accuracy of 'difficult-to-detect' landmarks by using spatial guidance from 'easy-to-detect' landmarks. To evaluate our method, we conducted experiments extensively on three datasets for detecting prostate landmarks and head & neck landmarks in computed tomography images, and also dental landmarks in cone beam computed tomography images. The results show the effectiveness of our collaborative landmark detection framework in improving landmark detection accuracy, compared to other state-of-the-art methods.

AB - Anatomical landmark detection plays an important role in medical image analysis, e.g. for registration, segmentation and quantitative analysis. Among the various existing methods for landmark detection, regression-based methods have recently attracted much attention due to their robustness and efficiency. In these methods, landmarks are localised through voting from all image voxels, which is completely different from the classification-based methods that use voxel-wise classification to detect landmarks. Despite their robustness, the accuracy of regression-based landmark detection methods is often limited due to (1) the inclusion of uninformative image voxels in the voting procedure, and (2) the lack of effective ways to incorporate inter-landmark spatial dependency into the detection step. In this paper, we propose a collaborative landmark detection framework to address these limitations. The concept of collaboration is reflected in two aspects. (1) Multi-resolution collaboration. A multi-resolution strategy is proposed to hierarchically localise landmarks by gradually excluding uninformative votes from faraway voxels. Moreover, for informative voxels near the landmark, a spherical sampling strategy is also designed at the training stage to improve their prediction accuracy. (2) Inter-landmark collaboration. A confidence-based landmark detection strategy is proposed to improve the detection accuracy of 'difficult-to-detect' landmarks by using spatial guidance from 'easy-to-detect' landmarks. To evaluate our method, we conducted experiments extensively on three datasets for detecting prostate landmarks and head & neck landmarks in computed tomography images, and also dental landmarks in cone beam computed tomography images. The results show the effectiveness of our collaborative landmark detection framework in improving landmark detection accuracy, compared to other state-of-the-art methods.

KW - image analysis

KW - image processing

KW - image registration

KW - landmark detection

KW - landmark-based analysis

KW - machine learning

KW - prostate localization

UR - http://www.scopus.com/inward/record.url?scp=84957922569&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84957922569&partnerID=8YFLogxK

U2 - 10.1088/0031-9155/60/24/9377

DO - 10.1088/0031-9155/60/24/9377

M3 - Article

C2 - 26579736

AN - SCOPUS:84957922569

VL - 60

SP - 9377

EP - 9401

JO - Physics in Medicine and Biology

JF - Physics in Medicine and Biology

SN - 0031-9155

IS - 24

ER -