Efficient brain lesion segmentation using multi-modality tissue-based feature selection and support vector machines.
Identifieur interne : 001092 ( Ncbi/Checkpoint ); précédent : 001091; suivant : 001093Efficient brain lesion segmentation using multi-modality tissue-based feature selection and support vector machines.
Auteurs : Jean-Baptiste Fiot [Australie] ; Laurent D. Cohen ; Parnesh Raniga ; Jurgen FrippSource :
- International journal for numerical methods in biomedical engineering [ 2040-7947 ] ; 2013.
Descripteurs français
- KwdFr :
- MESH :
- anatomie et histologie : Encéphale.
- anatomopathologie : Encéphale, Tumeurs du cerveau.
- Humains, Imagerie par résonance magnétique, Machine à vecteur de support, Neuroimagerie, Reproductibilité des résultats, Traitement d'image par ordinateur.
English descriptors
- KwdEn :
- MESH :
- anatomy & histology : Brain.
- methods : Image Processing, Computer-Assisted, Neuroimaging.
- pathology : Brain, Brain Neoplasms.
- Humans, Magnetic Resonance Imaging, Reproducibility of Results, Support Vector Machine.
Abstract
Support vector machines (SVM) are machine learning techniques that have been used for segmentation and classification of medical images, including segmentation of white matter hyper-intensities (WMH). Current approaches using SVM for WMH segmentation extract features from the brain and classify these followed by complex post-processing steps to remove false positives. The method presented in this paper combines advanced pre-processing, tissue-based feature selection and SVM classification to obtain efficient and accurate WMH segmentation. Features from 125 patients, generated from up to four MR modalities [T1-w, T2-w, proton-density and fluid attenuated inversion recovery(FLAIR)], differing neighbourhood sizes and the use of multi-scale features were compared. We found that although using all four modalities gave the best overall classification (average Dice scores of 0.54 ± 0.12, 0.72 ± 0.06 and 0.82 ± 0.06 respectively for small, moderate and severe lesion loads); this was not significantly different (p = 0.50) from using just T1-w and FLAIR sequences (Dice scores of 0.52 ± 0.13, 0.71 ± 0.08 and 0.81 ± 0.07). Furthermore, there was a negligible difference between using 5 × 5 × 5 and 3 × 3 × 3 features (p = 0.93). Finally, we show that careful consideration of features and pre-processing techniques not only saves storage space and computation time but also leads to more efficient classification, which outperforms the one based on all features with post-processing.
DOI: 10.1002/cnm.2537
PubMed: 23303595
Affiliations:
Links toward previous steps (curation, corpus...)
- to stream PubMed, to step Corpus: 003B30
- to stream PubMed, to step Curation: 003A00
- to stream PubMed, to step Checkpoint: 003A00
- to stream Ncbi, to step Merge: 001092
- to stream Ncbi, to step Curation: 001092
Links to Exploration step
pubmed:23303595Le document en format XML
<record><TEI><teiHeader><fileDesc><titleStmt><title xml:lang="en">Efficient brain lesion segmentation using multi-modality tissue-based feature selection and support vector machines.</title><author><name sortKey="Fiot, Jean Baptiste" sort="Fiot, Jean Baptiste" uniqKey="Fiot J" first="Jean-Baptiste" last="Fiot">Jean-Baptiste Fiot</name><affiliation wicri:level="1"><nlm:affiliation>CEREMADE, UMR 7534 CNRS Université Paris Dauphine, France; CSIRO Preventative Health National Research Flagship ICTC, The Australian e-Health Research Centre - BioMedIA, Royal Brisbane and Women's Hospital, Herston, Qld, Australia.</nlm:affiliation><country xml:lang="fr">Australie</country><wicri:regionArea>CEREMADE, UMR 7534 CNRS Université Paris Dauphine, France; CSIRO Preventative Health National Research Flagship ICTC, The Australian e-Health Research Centre - BioMedIA, Royal Brisbane and Women's Hospital, Herston, Qld</wicri:regionArea><wicri:noRegion>Qld</wicri:noRegion></affiliation></author><author><name sortKey="Cohen, Laurent D" sort="Cohen, Laurent D" uniqKey="Cohen L" first="Laurent D" last="Cohen">Laurent D. Cohen</name></author><author><name sortKey="Raniga, Parnesh" sort="Raniga, Parnesh" uniqKey="Raniga P" first="Parnesh" last="Raniga">Parnesh Raniga</name></author><author><name sortKey="Fripp, Jurgen" sort="Fripp, Jurgen" uniqKey="Fripp J" first="Jurgen" last="Fripp">Jurgen Fripp</name></author></titleStmt><publicationStmt><idno type="wicri:source">PubMed</idno><date when="2013">2013</date><idno type="RBID">pubmed:23303595</idno><idno type="pmid">23303595</idno><idno type="doi">10.1002/cnm.2537</idno><idno type="wicri:Area/PubMed/Corpus">003B30</idno><idno type="wicri:explorRef" wicri:stream="PubMed" wicri:step="Corpus" wicri:corpus="PubMed">003B30</idno><idno type="wicri:Area/PubMed/Curation">003A00</idno><idno type="wicri:explorRef" wicri:stream="PubMed" wicri:step="Curation">003A00</idno><idno type="wicri:Area/PubMed/Checkpoint">003A00</idno><idno type="wicri:explorRef" wicri:stream="Checkpoint" wicri:step="PubMed">003A00</idno><idno type="wicri:Area/Ncbi/Merge">001092</idno><idno type="wicri:Area/Ncbi/Curation">001092</idno><idno type="wicri:Area/Ncbi/Checkpoint">001092</idno></publicationStmt><sourceDesc><biblStruct><analytic><title xml:lang="en">Efficient brain lesion segmentation using multi-modality tissue-based feature selection and support vector machines.</title><author><name sortKey="Fiot, Jean Baptiste" sort="Fiot, Jean Baptiste" uniqKey="Fiot J" first="Jean-Baptiste" last="Fiot">Jean-Baptiste Fiot</name><affiliation wicri:level="1"><nlm:affiliation>CEREMADE, UMR 7534 CNRS Université Paris Dauphine, France; CSIRO Preventative Health National Research Flagship ICTC, The Australian e-Health Research Centre - BioMedIA, Royal Brisbane and Women's Hospital, Herston, Qld, Australia.</nlm:affiliation><country xml:lang="fr">Australie</country><wicri:regionArea>CEREMADE, UMR 7534 CNRS Université Paris Dauphine, France; CSIRO Preventative Health National Research Flagship ICTC, The Australian e-Health Research Centre - BioMedIA, Royal Brisbane and Women's Hospital, Herston, Qld</wicri:regionArea><wicri:noRegion>Qld</wicri:noRegion></affiliation></author><author><name sortKey="Cohen, Laurent D" sort="Cohen, Laurent D" uniqKey="Cohen L" first="Laurent D" last="Cohen">Laurent D. Cohen</name></author><author><name sortKey="Raniga, Parnesh" sort="Raniga, Parnesh" uniqKey="Raniga P" first="Parnesh" last="Raniga">Parnesh Raniga</name></author><author><name sortKey="Fripp, Jurgen" sort="Fripp, Jurgen" uniqKey="Fripp J" first="Jurgen" last="Fripp">Jurgen Fripp</name></author></analytic><series><title level="j">International journal for numerical methods in biomedical engineering</title><idno type="eISSN">2040-7947</idno><imprint><date when="2013" type="published">2013</date></imprint></series></biblStruct></sourceDesc></fileDesc><profileDesc><textClass><keywords scheme="KwdEn" xml:lang="en"><term>Brain (anatomy & histology)</term><term>Brain (pathology)</term><term>Brain Neoplasms (pathology)</term><term>Humans</term><term>Image Processing, Computer-Assisted (methods)</term><term>Magnetic Resonance Imaging</term><term>Neuroimaging (methods)</term><term>Reproducibility of Results</term><term>Support Vector Machine</term></keywords><keywords scheme="KwdFr" xml:lang="fr"><term>Encéphale (anatomie et histologie)</term><term>Encéphale (anatomopathologie)</term><term>Humains</term><term>Imagerie par résonance magnétique</term><term>Machine à vecteur de support</term><term>Neuroimagerie ()</term><term>Reproductibilité des résultats</term><term>Traitement d'image par ordinateur ()</term><term>Tumeurs du cerveau (anatomopathologie)</term></keywords><keywords scheme="MESH" qualifier="anatomie et histologie" xml:lang="fr"><term>Encéphale</term></keywords><keywords scheme="MESH" qualifier="anatomopathologie" xml:lang="fr"><term>Encéphale</term><term>Tumeurs du cerveau</term></keywords><keywords scheme="MESH" qualifier="anatomy & histology" xml:lang="en"><term>Brain</term></keywords><keywords scheme="MESH" qualifier="methods" xml:lang="en"><term>Image Processing, Computer-Assisted</term><term>Neuroimaging</term></keywords><keywords scheme="MESH" qualifier="pathology" xml:lang="en"><term>Brain</term><term>Brain Neoplasms</term></keywords><keywords scheme="MESH" xml:lang="en"><term>Humans</term><term>Magnetic Resonance Imaging</term><term>Reproducibility of Results</term><term>Support Vector Machine</term></keywords><keywords scheme="MESH" xml:lang="fr"><term>Humains</term><term>Imagerie par résonance magnétique</term><term>Machine à vecteur de support</term><term>Neuroimagerie</term><term>Reproductibilité des résultats</term><term>Traitement d'image par ordinateur</term></keywords></textClass></profileDesc></teiHeader><front><div type="abstract" xml:lang="en">Support vector machines (SVM) are machine learning techniques that have been used for segmentation and classification of medical images, including segmentation of white matter hyper-intensities (WMH). Current approaches using SVM for WMH segmentation extract features from the brain and classify these followed by complex post-processing steps to remove false positives. The method presented in this paper combines advanced pre-processing, tissue-based feature selection and SVM classification to obtain efficient and accurate WMH segmentation. Features from 125 patients, generated from up to four MR modalities [T1-w, T2-w, proton-density and fluid attenuated inversion recovery(FLAIR)], differing neighbourhood sizes and the use of multi-scale features were compared. We found that although using all four modalities gave the best overall classification (average Dice scores of 0.54 ± 0.12, 0.72 ± 0.06 and 0.82 ± 0.06 respectively for small, moderate and severe lesion loads); this was not significantly different (p = 0.50) from using just T1-w and FLAIR sequences (Dice scores of 0.52 ± 0.13, 0.71 ± 0.08 and 0.81 ± 0.07). Furthermore, there was a negligible difference between using 5 × 5 × 5 and 3 × 3 × 3 features (p = 0.93). Finally, we show that careful consideration of features and pre-processing techniques not only saves storage space and computation time but also leads to more efficient classification, which outperforms the one based on all features with post-processing.</div></front></TEI><affiliations><list><country><li>Australie</li></country></list><tree><noCountry><name sortKey="Cohen, Laurent D" sort="Cohen, Laurent D" uniqKey="Cohen L" first="Laurent D" last="Cohen">Laurent D. Cohen</name><name sortKey="Fripp, Jurgen" sort="Fripp, Jurgen" uniqKey="Fripp J" first="Jurgen" last="Fripp">Jurgen Fripp</name><name sortKey="Raniga, Parnesh" sort="Raniga, Parnesh" uniqKey="Raniga P" first="Parnesh" last="Raniga">Parnesh Raniga</name></noCountry><country name="Australie"><noRegion><name sortKey="Fiot, Jean Baptiste" sort="Fiot, Jean Baptiste" uniqKey="Fiot J" first="Jean-Baptiste" last="Fiot">Jean-Baptiste Fiot</name></noRegion></country></tree></affiliations></record>Pour manipuler ce document sous Unix (Dilib)
EXPLOR_STEP=$WICRI_ROOT/Wicri/Asie/explor/AustralieFrV1/Data/Ncbi/Checkpoint
HfdSelect -h $EXPLOR_STEP/biblio.hfd -nk 001092 | SxmlIndent | more
Ou
HfdSelect -h $EXPLOR_AREA/Data/Ncbi/Checkpoint/biblio.hfd -nk 001092 | SxmlIndent | more
Pour mettre un lien sur cette page dans le réseau Wicri
{{Explor lien
|wiki= Wicri/Asie
|area= AustralieFrV1
|flux= Ncbi
|étape= Checkpoint
|type= RBID
|clé= pubmed:23303595
|texte= Efficient brain lesion segmentation using multi-modality tissue-based feature selection and support vector machines.
}}
Pour générer des pages wiki
HfdIndexSelect -h $EXPLOR_AREA/Data/Ncbi/Checkpoint/RBID.i -Sk "pubmed:23303595" \
| HfdSelect -Kh $EXPLOR_AREA/Data/Ncbi/Checkpoint/biblio.hfd \
| NlmPubMed2Wicri -a AustralieFrV1
| This area was generated with Dilib version V0.6.33. |  |