Volumetric Image Registration From Invariant Keypoints

We present a method for image registration based on 3D scale- and rotation-invariant keypoints. The method extends the scale invariant feature transform (SIFT) to arbitrary dimensions by making key modifications to orientation assignment and gradient histograms. Rotation invariance is proven mathema...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on image processing 2017-10, Vol.26 (10), p.4900-4910
Main Authors: Rister, Blaine, Horowitz, Mark A., Rubin, Daniel L.
Format: Article
Language:English
Subjects:
3D SIFT
Affine transformations
Biomedical imaging
Brain
Computed tomography
computed tomography (CT)
Computer simulation
Computer vision
Head
Histograms
Image registration
Invariants
Lesions
Libraries
Magnetic resonance imaging
magnetic resonance imaging (MRI)
Mathematical analysis
medical image registration
Medical imaging
Multiple sclerosis
Registration
Tensile stress
Three-dimensional displays
Transformations (mathematics)
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
cited_by cdi_FETCH-LOGICAL-c444t-7650b30b9ca101e15ab8234343ca8f8f42e8623b8cf0cbee8ff3bf945096667c3
cites cdi_FETCH-LOGICAL-c444t-7650b30b9ca101e15ab8234343ca8f8f42e8623b8cf0cbee8ff3bf945096667c3
container_end_page 4910
container_issue 10
container_start_page 4900
container_title IEEE transactions on image processing
container_volume 26
creator Rister, Blaine
Horowitz, Mark A.
Rubin, Daniel L.
description We present a method for image registration based on 3D scale- and rotation-invariant keypoints. The method extends the scale invariant feature transform (SIFT) to arbitrary dimensions by making key modifications to orientation assignment and gradient histograms. Rotation invariance is proven mathematically. Additional modifications are made to extrema detection and keypoint matching based on the demands of image registration. Our experiments suggest that the choice of neighborhood in discrete extrema detection has a strong impact on image registration accuracy. In head MR images, the brain is registered to a labeled atlas with an average Dice coefficient of 92%, outperforming registration from mutual information as well as an existing 3D SIFT implementation. In abdominal CT images, the spine is registered with an average error of 4.82 mm. Furthermore, keypoints are matched with high precision in simulated head MR images exhibiting lesions from multiple sclerosis. These results were achieved using only affine transforms, and with no change in parameters across a wide variety of medical images. This paper is freely available as a cross-platform software library.
doi_str_mv 10.1109/TIP.2017.2722689
format article
fullrecord <record><control><sourceid>proquest_ieee_</sourceid><recordid>TN_cdi_ieee_primary_7967757</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>7967757</ieee_id><sourcerecordid>1916711348</sourcerecordid><originalsourceid>FETCH-LOGICAL-c444t-7650b30b9ca101e15ab8234343ca8f8f42e8623b8cf0cbee8ff3bf945096667c3</originalsourceid><addsrcrecordid>eNpdkU1r3DAQhkVoyVd7DxSCoZdcvNHI-rwUSkjapYGEkPYqZGW8VbCtjWQH8u-jsNulLTqMYJ55meEh5AToAoCa8_vl7YJRUAumGJPa7JFDMBxqSjl7V_5UqFoBNwfkKOdHSoELkPvkgGmpGRPykMhfsZ8HnFLw1XJwK6zucBXylNwU4lhdpThUy_HZpeDGqfqBL-sYxil_IO8712f8uK3H5OfV5f3F9_r65tvy4ut17TnnU62koG1DW-MdUEAQrtWs4eV5pzvdcYZasqbVvqO-RdRd17Sd4YIaKaXyzTH5ssldz-2ADx7Hsllv1ykMLr3Y6IL9tzOG33YVn60QGgSHEnC2DUjxacY82SFkj33vRoxztmBAKoCG64J-_g99jHMay3mFYo0SXEpeKLqhfIo5J-x2ywC1b1JskWLfpNitlDJy-vcRu4E_FgrwaQMERNy1lZFKCdW8Ak8DkE0</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>1923754664</pqid></control><display><type>article</type><title>Volumetric Image Registration From Invariant Keypoints</title><source>IEEE Electronic Library (IEL) Journals</source><creator>Rister, Blaine ; Horowitz, Mark A. ; Rubin, Daniel L.</creator><creatorcontrib>Rister, Blaine ; Horowitz, Mark A. ; Rubin, Daniel L.</creatorcontrib><description>We present a method for image registration based on 3D scale- and rotation-invariant keypoints. The method extends the scale invariant feature transform (SIFT) to arbitrary dimensions by making key modifications to orientation assignment and gradient histograms. Rotation invariance is proven mathematically. Additional modifications are made to extrema detection and keypoint matching based on the demands of image registration. Our experiments suggest that the choice of neighborhood in discrete extrema detection has a strong impact on image registration accuracy. In head MR images, the brain is registered to a labeled atlas with an average Dice coefficient of 92%, outperforming registration from mutual information as well as an existing 3D SIFT implementation. In abdominal CT images, the spine is registered with an average error of 4.82 mm. Furthermore, keypoints are matched with high precision in simulated head MR images exhibiting lesions from multiple sclerosis. These results were achieved using only affine transforms, and with no change in parameters across a wide variety of medical images. This paper is freely available as a cross-platform software library.</description><identifier>ISSN: 1057-7149</identifier><identifier>EISSN: 1941-0042</identifier><identifier>DOI: 10.1109/TIP.2017.2722689</identifier><identifier>PMID: 28682256</identifier><identifier>CODEN: IIPRE4</identifier><language>eng</language><publisher>United States: IEEE</publisher><subject>3D SIFT ; Affine transformations ; Biomedical imaging ; Brain ; Computed tomography ; computed tomography (CT) ; Computer simulation ; Computer vision ; Head ; Histograms ; Image registration ; Invariants ; Lesions ; Libraries ; Magnetic resonance imaging ; magnetic resonance imaging (MRI) ; Mathematical analysis ; medical image registration ; Medical imaging ; Multiple sclerosis ; Registration ; Tensile stress ; Three-dimensional displays ; Transformations (mathematics)</subject><ispartof>IEEE transactions on image processing, 2017-10, Vol.26 (10), p.4900-4910</ispartof><rights>Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2017</rights><lds50>peer_reviewed</lds50><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c444t-7650b30b9ca101e15ab8234343ca8f8f42e8623b8cf0cbee8ff3bf945096667c3</citedby><cites>FETCH-LOGICAL-c444t-7650b30b9ca101e15ab8234343ca8f8f42e8623b8cf0cbee8ff3bf945096667c3</cites><orcidid>0000-0002-4490-0444</orcidid></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://ieeexplore.ieee.org/document/7967757$$EHTML$$P50$$Gieee$$H</linktohtml><link.rule.ids>230,315,786,790,891,27957,27958,55147</link.rule.ids><backlink>$$Uhttps://www.ncbi.nlm.nih.gov/pubmed/28682256$$D View this record in MEDLINE/PubMed$$Hfree_for_read</backlink></links><search><creatorcontrib>Rister, Blaine</creatorcontrib><creatorcontrib>Horowitz, Mark A.</creatorcontrib><creatorcontrib>Rubin, Daniel L.</creatorcontrib><title>Volumetric Image Registration From Invariant Keypoints</title><title>IEEE transactions on image processing</title><addtitle>TIP</addtitle><addtitle>IEEE Trans Image Process</addtitle><description>We present a method for image registration based on 3D scale- and rotation-invariant keypoints. The method extends the scale invariant feature transform (SIFT) to arbitrary dimensions by making key modifications to orientation assignment and gradient histograms. Rotation invariance is proven mathematically. Additional modifications are made to extrema detection and keypoint matching based on the demands of image registration. Our experiments suggest that the choice of neighborhood in discrete extrema detection has a strong impact on image registration accuracy. In head MR images, the brain is registered to a labeled atlas with an average Dice coefficient of 92%, outperforming registration from mutual information as well as an existing 3D SIFT implementation. In abdominal CT images, the spine is registered with an average error of 4.82 mm. Furthermore, keypoints are matched with high precision in simulated head MR images exhibiting lesions from multiple sclerosis. These results were achieved using only affine transforms, and with no change in parameters across a wide variety of medical images. This paper is freely available as a cross-platform software library.</description><subject>3D SIFT</subject><subject>Affine transformations</subject><subject>Biomedical imaging</subject><subject>Brain</subject><subject>Computed tomography</subject><subject>computed tomography (CT)</subject><subject>Computer simulation</subject><subject>Computer vision</subject><subject>Head</subject><subject>Histograms</subject><subject>Image registration</subject><subject>Invariants</subject><subject>Lesions</subject><subject>Libraries</subject><subject>Magnetic resonance imaging</subject><subject>magnetic resonance imaging (MRI)</subject><subject>Mathematical analysis</subject><subject>medical image registration</subject><subject>Medical imaging</subject><subject>Multiple sclerosis</subject><subject>Registration</subject><subject>Tensile stress</subject><subject>Three-dimensional displays</subject><subject>Transformations (mathematics)</subject><issn>1057-7149</issn><issn>1941-0042</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2017</creationdate><recordtype>article</recordtype><recordid>eNpdkU1r3DAQhkVoyVd7DxSCoZdcvNHI-rwUSkjapYGEkPYqZGW8VbCtjWQH8u-jsNulLTqMYJ55meEh5AToAoCa8_vl7YJRUAumGJPa7JFDMBxqSjl7V_5UqFoBNwfkKOdHSoELkPvkgGmpGRPykMhfsZ8HnFLw1XJwK6zucBXylNwU4lhdpThUy_HZpeDGqfqBL-sYxil_IO8712f8uK3H5OfV5f3F9_r65tvy4ut17TnnU62koG1DW-MdUEAQrtWs4eV5pzvdcYZasqbVvqO-RdRd17Sd4YIaKaXyzTH5ssldz-2ADx7Hsllv1ykMLr3Y6IL9tzOG33YVn60QGgSHEnC2DUjxacY82SFkj33vRoxztmBAKoCG64J-_g99jHMay3mFYo0SXEpeKLqhfIo5J-x2ywC1b1JskWLfpNitlDJy-vcRu4E_FgrwaQMERNy1lZFKCdW8Ak8DkE0</recordid><startdate>20171001</startdate><enddate>20171001</enddate><creator>Rister, Blaine</creator><creator>Horowitz, Mark A.</creator><creator>Rubin, Daniel L.</creator><general>IEEE</general><general>The Institute of Electrical and Electronics Engineers, Inc. (IEEE)</general><scope>97E</scope><scope>RIA</scope><scope>RIE</scope><scope>NPM</scope><scope>AAYXX</scope><scope>CITATION</scope><scope>7SC</scope><scope>7SP</scope><scope>8FD</scope><scope>JQ2</scope><scope>L7M</scope><scope>L~C</scope><scope>L~D</scope><scope>7X8</scope><scope>5PM</scope><orcidid>https://orcid.org/0000-0002-4490-0444</orcidid></search><sort><creationdate>20171001</creationdate><title>Volumetric Image Registration From Invariant Keypoints</title><author>Rister, Blaine ; Horowitz, Mark A. ; Rubin, Daniel L.</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c444t-7650b30b9ca101e15ab8234343ca8f8f42e8623b8cf0cbee8ff3bf945096667c3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2017</creationdate><topic>3D SIFT</topic><topic>Affine transformations</topic><topic>Biomedical imaging</topic><topic>Brain</topic><topic>Computed tomography</topic><topic>computed tomography (CT)</topic><topic>Computer simulation</topic><topic>Computer vision</topic><topic>Head</topic><topic>Histograms</topic><topic>Image registration</topic><topic>Invariants</topic><topic>Lesions</topic><topic>Libraries</topic><topic>Magnetic resonance imaging</topic><topic>magnetic resonance imaging (MRI)</topic><topic>Mathematical analysis</topic><topic>medical image registration</topic><topic>Medical imaging</topic><topic>Multiple sclerosis</topic><topic>Registration</topic><topic>Tensile stress</topic><topic>Three-dimensional displays</topic><topic>Transformations (mathematics)</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Rister, Blaine</creatorcontrib><creatorcontrib>Horowitz, Mark A.</creatorcontrib><creatorcontrib>Rubin, Daniel L.</creatorcontrib><collection>IEEE All-Society Periodicals Package (ASPP) 2005-present</collection><collection>IEEE All-Society Periodicals Package (ASPP) 1998–Present</collection><collection>IEEE Electronic Library (IEL)</collection><collection>PubMed</collection><collection>CrossRef</collection><collection>Computer and Information Systems Abstracts</collection><collection>Electronics &amp; Communications Abstracts</collection><collection>Technology Research Database</collection><collection>ProQuest Computer Science Collection</collection><collection>Advanced Technologies Database with Aerospace</collection><collection>Computer and Information Systems Abstracts – Academic</collection><collection>Computer and Information Systems Abstracts Professional</collection><collection>MEDLINE - Academic</collection><collection>PubMed Central (Full Participant titles)</collection><jtitle>IEEE transactions on image processing</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Rister, Blaine</au><au>Horowitz, Mark A.</au><au>Rubin, Daniel L.</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Volumetric Image Registration From Invariant Keypoints</atitle><jtitle>IEEE transactions on image processing</jtitle><stitle>TIP</stitle><addtitle>IEEE Trans Image Process</addtitle><date>2017-10-01</date><risdate>2017</risdate><volume>26</volume><issue>10</issue><spage>4900</spage><epage>4910</epage><pages>4900-4910</pages><issn>1057-7149</issn><eissn>1941-0042</eissn><coden>IIPRE4</coden><notes>ObjectType-Article-1</notes><notes>SourceType-Scholarly Journals-1</notes><notes>ObjectType-Feature-2</notes><notes>content type line 23</notes><abstract>We present a method for image registration based on 3D scale- and rotation-invariant keypoints. The method extends the scale invariant feature transform (SIFT) to arbitrary dimensions by making key modifications to orientation assignment and gradient histograms. Rotation invariance is proven mathematically. Additional modifications are made to extrema detection and keypoint matching based on the demands of image registration. Our experiments suggest that the choice of neighborhood in discrete extrema detection has a strong impact on image registration accuracy. In head MR images, the brain is registered to a labeled atlas with an average Dice coefficient of 92%, outperforming registration from mutual information as well as an existing 3D SIFT implementation. In abdominal CT images, the spine is registered with an average error of 4.82 mm. Furthermore, keypoints are matched with high precision in simulated head MR images exhibiting lesions from multiple sclerosis. These results were achieved using only affine transforms, and with no change in parameters across a wide variety of medical images. This paper is freely available as a cross-platform software library.</abstract><cop>United States</cop><pub>IEEE</pub><pmid>28682256</pmid><doi>10.1109/TIP.2017.2722689</doi><tpages>11</tpages><orcidid>https://orcid.org/0000-0002-4490-0444</orcidid><oa>free_for_read</oa></addata></record>
fulltext fulltext
identifier ISSN: 1057-7149
ispartof IEEE transactions on image processing, 2017-10, Vol.26 (10), p.4900-4910
issn 1057-7149
1941-0042
language eng
recordid cdi_ieee_primary_7967757
source IEEE Electronic Library (IEL) Journals
subjects 3D SIFT
Affine transformations
Biomedical imaging
Brain
Computed tomography
computed tomography (CT)
Computer simulation
Computer vision
Head
Histograms
Image registration
Invariants
Lesions
Libraries
Magnetic resonance imaging
magnetic resonance imaging (MRI)
Mathematical analysis
medical image registration
Medical imaging
Multiple sclerosis
Registration
Tensile stress
Three-dimensional displays
Transformations (mathematics)
title Volumetric Image Registration From Invariant Keypoints
url http://sfxeu10.hosted.exlibrisgroup.com/loughborough?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-09-21T20%3A23%3A07IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_ieee_&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Volumetric%20Image%20Registration%20From%20Invariant%20Keypoints&rft.jtitle=IEEE%20transactions%20on%20image%20processing&rft.au=Rister,%20Blaine&rft.date=2017-10-01&rft.volume=26&rft.issue=10&rft.spage=4900&rft.epage=4910&rft.pages=4900-4910&rft.issn=1057-7149&rft.eissn=1941-0042&rft.coden=IIPRE4&rft_id=info:doi/10.1109/TIP.2017.2722689&rft_dat=%3Cproquest_ieee_%3E1916711348%3C/proquest_ieee_%3E%3Cgrp_id%3Ecdi_FETCH-LOGICAL-c444t-7650b30b9ca101e15ab8234343ca8f8f42e8623b8cf0cbee8ff3bf945096667c3%3C/grp_id%3E%3Coa%3E%3C/oa%3E%3Curl%3E%3C/url%3E&rft_id=info:oai/&rft_pqid=1923754664&rft_id=info:pmid/28682256&rft_ieee_id=7967757&rfr_iscdi=true