Marginal Space Learning for Efficient Detection of 2D/3D Anatomical Structures in Medical Images View Full Text


Ontology type: schema:Chapter     


Chapter Info

DATE

2009

AUTHORS

Yefeng Zheng , Bogdan Georgescu , Dorin Comaniciu

ABSTRACT

Recently, marginal space learning (MSL) was proposed as a generic approach for automatic detection of 3D anatomical structures in many medical imaging modalities [1]. To accurately localize a 3D object, we need to estimate nine pose parameters (three for position, three for orientation, and three for anisotropic scaling). Instead of exhaustively searching the original nine-dimen-sional pose parameter space, only low-dimensional marginal spaces are searched in MSL to improve the detection speed. In this paper, we apply MSL to 2D object detection and perform a thorough comparison between MSL and the alternative full space learning (FSL) approach. Experiments on left ventricle detection in 2D MRI images show MSL outperforms FSL in both speed and accuracy. In addition, we propose two novel techniques, constrained MSL and nonrigid MSL, to further improve the efficiency and accuracy. In many real applications, a strong correlation may exist among pose parameters in the same marginal spaces. For example, a large object may have large scaling values along all directions. Constrained MSL exploits this correlation for further speed-up. The original MSL only estimates the rigid transformation of an object in the image, therefore cannot accurately localize a nonrigid object under a large deformation. The proposed nonrigid MSL directly estimates the nonrigid deformation parameters to improve the localization accuracy. The comparison experiments on liver detection in 226 abdominal CT volumes demonstrate the effectiveness of the proposed methods. Our system takes less than a second to accurately detect the liver in a volume. More... »

PAGES

411-422

Identifiers

URI

http://scigraph.springernature.com/pub.10.1007/978-3-642-02498-6_34

DOI

http://dx.doi.org/10.1007/978-3-642-02498-6_34

DIMENSIONS

https://app.dimensions.ai/details/publication/pub.1034632087

PUBMED

https://www.ncbi.nlm.nih.gov/pubmed/19694281


Indexing Status Check whether this publication has been indexed by Scopus and Web Of Science using the SN Indexing Status Tool
Incoming Citations Browse incoming citations for this publication using opencitations.net

JSON-LD is the canonical representation for SciGraph data.

TIP: You can open this SciGraph record using an external JSON-LD service: JSON-LD Playground Google SDTT

[
  {
    "@context": "https://springernature.github.io/scigraph/jsonld/sgcontext.json", 
    "about": [
      {
        "id": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/08", 
        "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
        "name": "Information and Computing Sciences", 
        "type": "DefinedTerm"
      }, 
      {
        "id": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/0801", 
        "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
        "name": "Artificial Intelligence and Image Processing", 
        "type": "DefinedTerm"
      }, 
      {
        "inDefinedTermSet": "https://www.nlm.nih.gov/mesh/", 
        "name": "Algorithms", 
        "type": "DefinedTerm"
      }, 
      {
        "inDefinedTermSet": "https://www.nlm.nih.gov/mesh/", 
        "name": "Artificial Intelligence", 
        "type": "DefinedTerm"
      }, 
      {
        "inDefinedTermSet": "https://www.nlm.nih.gov/mesh/", 
        "name": "Image Enhancement", 
        "type": "DefinedTerm"
      }, 
      {
        "inDefinedTermSet": "https://www.nlm.nih.gov/mesh/", 
        "name": "Image Interpretation, Computer-Assisted", 
        "type": "DefinedTerm"
      }, 
      {
        "inDefinedTermSet": "https://www.nlm.nih.gov/mesh/", 
        "name": "Imaging, Three-Dimensional", 
        "type": "DefinedTerm"
      }, 
      {
        "inDefinedTermSet": "https://www.nlm.nih.gov/mesh/", 
        "name": "Pattern Recognition, Automated", 
        "type": "DefinedTerm"
      }, 
      {
        "inDefinedTermSet": "https://www.nlm.nih.gov/mesh/", 
        "name": "Reproducibility of Results", 
        "type": "DefinedTerm"
      }, 
      {
        "inDefinedTermSet": "https://www.nlm.nih.gov/mesh/", 
        "name": "Sensitivity and Specificity", 
        "type": "DefinedTerm"
      }
    ], 
    "author": [
      {
        "affiliation": {
          "alternateName": "Integrated Data Systems Department, Siemens Corporate Research, USA", 
          "id": "http://www.grid.ac/institutes/grid.419233.e", 
          "name": [
            "Integrated Data Systems Department, Siemens Corporate Research, USA"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Zheng", 
        "givenName": "Yefeng", 
        "id": "sg:person.0767211426.21", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.0767211426.21"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "Integrated Data Systems Department, Siemens Corporate Research, USA", 
          "id": "http://www.grid.ac/institutes/grid.419233.e", 
          "name": [
            "Integrated Data Systems Department, Siemens Corporate Research, USA"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Georgescu", 
        "givenName": "Bogdan", 
        "id": "sg:person.0703547214.37", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.0703547214.37"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "Integrated Data Systems Department, Siemens Corporate Research, USA", 
          "id": "http://www.grid.ac/institutes/grid.419233.e", 
          "name": [
            "Integrated Data Systems Department, Siemens Corporate Research, USA"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Comaniciu", 
        "givenName": "Dorin", 
        "id": "sg:person.01066111014.77", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.01066111014.77"
        ], 
        "type": "Person"
      }
    ], 
    "datePublished": "2009", 
    "datePublishedReg": "2009-01-01", 
    "description": "Recently, marginal space learning (MSL) was proposed as a generic approach for automatic detection of 3D anatomical structures in many medical imaging modalities [1]. To accurately localize a 3D object, we need to estimate nine pose parameters (three for position, three for orientation, and three for anisotropic scaling). Instead of exhaustively searching the original nine-dimen-sional pose parameter space, only low-dimensional marginal spaces are searched in MSL to improve the detection speed. In this paper, we apply MSL to 2D object detection and perform a thorough comparison between MSL and the alternative full space learning (FSL) approach. Experiments on left ventricle detection in 2D MRI images show MSL outperforms FSL in both speed and accuracy. In addition, we propose two novel techniques, constrained MSL and nonrigid MSL, to further improve the efficiency and accuracy. In many real applications, a strong correlation may exist among pose parameters in the same marginal spaces. For example, a large object may have large scaling values along all directions. Constrained MSL exploits this correlation for further speed-up. The original MSL only estimates the rigid transformation of an object in the image, therefore cannot accurately localize a nonrigid object under a large deformation. The proposed nonrigid MSL directly estimates the nonrigid deformation parameters to improve the localization accuracy. The comparison experiments on liver detection in 226 abdominal CT volumes demonstrate the effectiveness of the proposed methods. Our system takes less than a second to accurately detect the liver in a volume.", 
    "editor": [
      {
        "familyName": "Prince", 
        "givenName": "Jerry L.", 
        "type": "Person"
      }, 
      {
        "familyName": "Pham", 
        "givenName": "Dzung L.", 
        "type": "Person"
      }, 
      {
        "familyName": "Myers", 
        "givenName": "Kyle J.", 
        "type": "Person"
      }
    ], 
    "genre": "chapter", 
    "id": "sg:pub.10.1007/978-3-642-02498-6_34", 
    "inLanguage": "en", 
    "isAccessibleForFree": false, 
    "isPartOf": {
      "isbn": [
        "978-3-642-02497-9", 
        "978-3-642-02498-6"
      ], 
      "name": "Information Processing in Medical Imaging", 
      "type": "Book"
    }, 
    "keywords": [
      "marginal space learning", 
      "space learning", 
      "pose parameters", 
      "abdominal CT volumes", 
      "pose parameter space", 
      "medical imaging modalities", 
      "medical images", 
      "ventricle detection", 
      "automatic detection", 
      "nonrigid objects", 
      "learning approach", 
      "detection speed", 
      "liver detection", 
      "real applications", 
      "localization accuracy", 
      "generic approach", 
      "rigid transformation", 
      "nonrigid deformations", 
      "CT volumes", 
      "anatomical structures", 
      "MRI images", 
      "objects", 
      "large objects", 
      "efficient detection", 
      "images", 
      "thorough comparison", 
      "novel technique", 
      "marginal spaces", 
      "learning", 
      "scaling value", 
      "accuracy", 
      "detection", 
      "space", 
      "speed", 
      "applications", 
      "parameter space", 
      "effectiveness", 
      "system", 
      "imaging modalities", 
      "technique", 
      "efficiency", 
      "example", 
      "FSL", 
      "method", 
      "experiments", 
      "parameters", 
      "transformation", 
      "structure", 
      "comparison", 
      "direction", 
      "volume", 
      "modalities", 
      "large deformation", 
      "addition", 
      "correlation", 
      "values", 
      "strong correlation", 
      "deformation", 
      "approach", 
      "paper", 
      "liver", 
      "nine-dimen-sional pose parameter space", 
      "low-dimensional marginal spaces", 
      "alternative full space learning (FSL) approach", 
      "full space learning (FSL) approach", 
      "space learning (FSL) approach", 
      "nonrigid MSL", 
      "same marginal spaces", 
      "large scaling values", 
      "Constrained MSL", 
      "original MSL"
    ], 
    "name": "Marginal Space Learning for Efficient Detection of 2D/3D Anatomical Structures in Medical Images", 
    "pagination": "411-422", 
    "productId": [
      {
        "name": "dimensions_id", 
        "type": "PropertyValue", 
        "value": [
          "pub.1034632087"
        ]
      }, 
      {
        "name": "doi", 
        "type": "PropertyValue", 
        "value": [
          "10.1007/978-3-642-02498-6_34"
        ]
      }, 
      {
        "name": "pubmed_id", 
        "type": "PropertyValue", 
        "value": [
          "19694281"
        ]
      }
    ], 
    "publisher": {
      "name": "Springer Nature", 
      "type": "Organisation"
    }, 
    "sameAs": [
      "https://doi.org/10.1007/978-3-642-02498-6_34", 
      "https://app.dimensions.ai/details/publication/pub.1034632087"
    ], 
    "sdDataset": "chapters", 
    "sdDatePublished": "2022-01-01T19:22", 
    "sdLicense": "https://scigraph.springernature.com/explorer/license/", 
    "sdPublisher": {
      "name": "Springer Nature - SN SciGraph project", 
      "type": "Organization"
    }, 
    "sdSource": "s3://com-springernature-scigraph/baseset/20220101/entities/gbq_results/chapter/chapter_390.jsonl", 
    "type": "Chapter", 
    "url": "https://doi.org/10.1007/978-3-642-02498-6_34"
  }
]
 

Download the RDF metadata as:  json-ld nt turtle xml License info

HOW TO GET THIS DATA PROGRAMMATICALLY:

JSON-LD is a popular format for linked data which is fully compatible with JSON.

curl -H 'Accept: application/ld+json' 'https://scigraph.springernature.com/pub.10.1007/978-3-642-02498-6_34'

N-Triples is a line-based linked data format ideal for batch operations.

curl -H 'Accept: application/n-triples' 'https://scigraph.springernature.com/pub.10.1007/978-3-642-02498-6_34'

Turtle is a human-readable linked data format.

curl -H 'Accept: text/turtle' 'https://scigraph.springernature.com/pub.10.1007/978-3-642-02498-6_34'

RDF/XML is a standard XML format for linked data.

curl -H 'Accept: application/rdf+xml' 'https://scigraph.springernature.com/pub.10.1007/978-3-642-02498-6_34'


 

This table displays all metadata directly associated to this object as RDF triples.

191 TRIPLES      23 PREDICATES      106 URIs      99 LITERALS      16 BLANK NODES

Subject Predicate Object
1 sg:pub.10.1007/978-3-642-02498-6_34 schema:about N26a50e3cf05e47ba8d20cb49283970bd
2 N441d267a13d743b29627192490e9279e
3 N50de0da9074d4098a6ea77973e1788ca
4 N54c341df7b9549cf97dfadf4fe05df11
5 N5aa0254ce3d7428a8e637f9f4d1e0d63
6 N81a5cd3983464ca28a7852678817afe6
7 Ndd1bc6ceb69540d2865468a705a5b9a7
8 Nfdfe149976d34c6ebc183cbb2a253489
9 anzsrc-for:08
10 anzsrc-for:0801
11 schema:author N30057e3d6e464612a0accb4f2bbf80fe
12 schema:datePublished 2009
13 schema:datePublishedReg 2009-01-01
14 schema:description Recently, marginal space learning (MSL) was proposed as a generic approach for automatic detection of 3D anatomical structures in many medical imaging modalities [1]. To accurately localize a 3D object, we need to estimate nine pose parameters (three for position, three for orientation, and three for anisotropic scaling). Instead of exhaustively searching the original nine-dimen-sional pose parameter space, only low-dimensional marginal spaces are searched in MSL to improve the detection speed. In this paper, we apply MSL to 2D object detection and perform a thorough comparison between MSL and the alternative full space learning (FSL) approach. Experiments on left ventricle detection in 2D MRI images show MSL outperforms FSL in both speed and accuracy. In addition, we propose two novel techniques, constrained MSL and nonrigid MSL, to further improve the efficiency and accuracy. In many real applications, a strong correlation may exist among pose parameters in the same marginal spaces. For example, a large object may have large scaling values along all directions. Constrained MSL exploits this correlation for further speed-up. The original MSL only estimates the rigid transformation of an object in the image, therefore cannot accurately localize a nonrigid object under a large deformation. The proposed nonrigid MSL directly estimates the nonrigid deformation parameters to improve the localization accuracy. The comparison experiments on liver detection in 226 abdominal CT volumes demonstrate the effectiveness of the proposed methods. Our system takes less than a second to accurately detect the liver in a volume.
15 schema:editor Naaa2393a72f04d11a429b102ce4a9212
16 schema:genre chapter
17 schema:inLanguage en
18 schema:isAccessibleForFree false
19 schema:isPartOf N40c33e425e414a37bcfbae4300d634bd
20 schema:keywords CT volumes
21 Constrained MSL
22 FSL
23 MRI images
24 abdominal CT volumes
25 accuracy
26 addition
27 alternative full space learning (FSL) approach
28 anatomical structures
29 applications
30 approach
31 automatic detection
32 comparison
33 correlation
34 deformation
35 detection
36 detection speed
37 direction
38 effectiveness
39 efficiency
40 efficient detection
41 example
42 experiments
43 full space learning (FSL) approach
44 generic approach
45 images
46 imaging modalities
47 large deformation
48 large objects
49 large scaling values
50 learning
51 learning approach
52 liver
53 liver detection
54 localization accuracy
55 low-dimensional marginal spaces
56 marginal space learning
57 marginal spaces
58 medical images
59 medical imaging modalities
60 method
61 modalities
62 nine-dimen-sional pose parameter space
63 nonrigid MSL
64 nonrigid deformations
65 nonrigid objects
66 novel technique
67 objects
68 original MSL
69 paper
70 parameter space
71 parameters
72 pose parameter space
73 pose parameters
74 real applications
75 rigid transformation
76 same marginal spaces
77 scaling value
78 space
79 space learning
80 space learning (FSL) approach
81 speed
82 strong correlation
83 structure
84 system
85 technique
86 thorough comparison
87 transformation
88 values
89 ventricle detection
90 volume
91 schema:name Marginal Space Learning for Efficient Detection of 2D/3D Anatomical Structures in Medical Images
92 schema:pagination 411-422
93 schema:productId N909efe11bd3b417494bcefd9e343422e
94 N9744e19ab8384ab08f627bc36a1e8d3d
95 Nb7f29790e6e04ad7a207440246079244
96 schema:publisher N31ea51c0a9514d61b8684c4e7cdb745b
97 schema:sameAs https://app.dimensions.ai/details/publication/pub.1034632087
98 https://doi.org/10.1007/978-3-642-02498-6_34
99 schema:sdDatePublished 2022-01-01T19:22
100 schema:sdLicense https://scigraph.springernature.com/explorer/license/
101 schema:sdPublisher N043b978c88694495817a0a6b54a56f66
102 schema:url https://doi.org/10.1007/978-3-642-02498-6_34
103 sgo:license sg:explorer/license/
104 sgo:sdDataset chapters
105 rdf:type schema:Chapter
106 N004ba9e4c2f142c38f98ff4a5273d3bf rdf:first N4c07eb7111644c0b8d151984698f5487
107 rdf:rest Ncc873f5fa4f049a9b5771c660fc5735b
108 N043b978c88694495817a0a6b54a56f66 schema:name Springer Nature - SN SciGraph project
109 rdf:type schema:Organization
110 N244781095f3846fe96f2dff9f435f5ef schema:familyName Myers
111 schema:givenName Kyle J.
112 rdf:type schema:Person
113 N26a50e3cf05e47ba8d20cb49283970bd schema:inDefinedTermSet https://www.nlm.nih.gov/mesh/
114 schema:name Artificial Intelligence
115 rdf:type schema:DefinedTerm
116 N30057e3d6e464612a0accb4f2bbf80fe rdf:first sg:person.0767211426.21
117 rdf:rest N5a33b318adea446c9f8ebd85b95c676f
118 N31ea51c0a9514d61b8684c4e7cdb745b schema:name Springer Nature
119 rdf:type schema:Organisation
120 N40c33e425e414a37bcfbae4300d634bd schema:isbn 978-3-642-02497-9
121 978-3-642-02498-6
122 schema:name Information Processing in Medical Imaging
123 rdf:type schema:Book
124 N441d267a13d743b29627192490e9279e schema:inDefinedTermSet https://www.nlm.nih.gov/mesh/
125 schema:name Image Interpretation, Computer-Assisted
126 rdf:type schema:DefinedTerm
127 N4c07eb7111644c0b8d151984698f5487 schema:familyName Pham
128 schema:givenName Dzung L.
129 rdf:type schema:Person
130 N50de0da9074d4098a6ea77973e1788ca schema:inDefinedTermSet https://www.nlm.nih.gov/mesh/
131 schema:name Pattern Recognition, Automated
132 rdf:type schema:DefinedTerm
133 N54c341df7b9549cf97dfadf4fe05df11 schema:inDefinedTermSet https://www.nlm.nih.gov/mesh/
134 schema:name Algorithms
135 rdf:type schema:DefinedTerm
136 N5a33b318adea446c9f8ebd85b95c676f rdf:first sg:person.0703547214.37
137 rdf:rest N8c3bf762734d4f5b8c68822286edf164
138 N5aa0254ce3d7428a8e637f9f4d1e0d63 schema:inDefinedTermSet https://www.nlm.nih.gov/mesh/
139 schema:name Image Enhancement
140 rdf:type schema:DefinedTerm
141 N81a5cd3983464ca28a7852678817afe6 schema:inDefinedTermSet https://www.nlm.nih.gov/mesh/
142 schema:name Imaging, Three-Dimensional
143 rdf:type schema:DefinedTerm
144 N8c3bf762734d4f5b8c68822286edf164 rdf:first sg:person.01066111014.77
145 rdf:rest rdf:nil
146 N909efe11bd3b417494bcefd9e343422e schema:name dimensions_id
147 schema:value pub.1034632087
148 rdf:type schema:PropertyValue
149 N9744e19ab8384ab08f627bc36a1e8d3d schema:name doi
150 schema:value 10.1007/978-3-642-02498-6_34
151 rdf:type schema:PropertyValue
152 Na1d6a99b25c945d2a44944bb9fadd2e9 schema:familyName Prince
153 schema:givenName Jerry L.
154 rdf:type schema:Person
155 Naaa2393a72f04d11a429b102ce4a9212 rdf:first Na1d6a99b25c945d2a44944bb9fadd2e9
156 rdf:rest N004ba9e4c2f142c38f98ff4a5273d3bf
157 Nb7f29790e6e04ad7a207440246079244 schema:name pubmed_id
158 schema:value 19694281
159 rdf:type schema:PropertyValue
160 Ncc873f5fa4f049a9b5771c660fc5735b rdf:first N244781095f3846fe96f2dff9f435f5ef
161 rdf:rest rdf:nil
162 Ndd1bc6ceb69540d2865468a705a5b9a7 schema:inDefinedTermSet https://www.nlm.nih.gov/mesh/
163 schema:name Reproducibility of Results
164 rdf:type schema:DefinedTerm
165 Nfdfe149976d34c6ebc183cbb2a253489 schema:inDefinedTermSet https://www.nlm.nih.gov/mesh/
166 schema:name Sensitivity and Specificity
167 rdf:type schema:DefinedTerm
168 anzsrc-for:08 schema:inDefinedTermSet anzsrc-for:
169 schema:name Information and Computing Sciences
170 rdf:type schema:DefinedTerm
171 anzsrc-for:0801 schema:inDefinedTermSet anzsrc-for:
172 schema:name Artificial Intelligence and Image Processing
173 rdf:type schema:DefinedTerm
174 sg:person.01066111014.77 schema:affiliation grid-institutes:grid.419233.e
175 schema:familyName Comaniciu
176 schema:givenName Dorin
177 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.01066111014.77
178 rdf:type schema:Person
179 sg:person.0703547214.37 schema:affiliation grid-institutes:grid.419233.e
180 schema:familyName Georgescu
181 schema:givenName Bogdan
182 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.0703547214.37
183 rdf:type schema:Person
184 sg:person.0767211426.21 schema:affiliation grid-institutes:grid.419233.e
185 schema:familyName Zheng
186 schema:givenName Yefeng
187 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.0767211426.21
188 rdf:type schema:Person
189 grid-institutes:grid.419233.e schema:alternateName Integrated Data Systems Department, Siemens Corporate Research, USA
190 schema:name Integrated Data Systems Department, Siemens Corporate Research, USA
191 rdf:type schema:Organization
 




Preview window. Press ESC to close (or click here)


...