Robust Detection of Multiple Sclerosis Lesions from Intensity-Normalized Multi-Channel MRI

Yogesh Karpate 1, * Olivier Commowick 1 Christian Barillot 1
* Auteur correspondant
1 VisAGeS - Vision, Action et Gestion d'informations <3ic/Export_liste_publi_HAL" v icbaril_k-g" /:2down">INSERMmat tiout N" al la noforch/index/q/*/stru
"col-md-12"> "col-md 1 trong>lyphicon-white"> Connexio
noforch/index/q/*/str/q/*/authFullName_s/Yogesh+Karpate" target="_blank">Yogesh Karpate 1, * noforch/index/q/*/str 1 / trong> ="nav nink5a31e14aa6bcc v> js/utile.js?3_20171120"> cryptM <("", nink5a31e14aa6bcc , lyphicon-'hite"> Connenvelope' ")ml { pos 1 noforch/ "col-md 1 trong>lyphicon-white"> Connexio
noforch/index/q/*/str/q/*/authFullName_s/Yogesh+Karpate" target="_blank">Olivier Commowick 1 noforch/index/q/*/str 1 / trong> 1 br/r< labe labe kplayDiv
ALr: /'_blank'" >R target="_bla ScientifV ss" m> target="_blv> >iv> 1 noforch/ "col-md 1 trong>lyphicon-white"> Connexio
noforc>1 /q/*/authIdHal_s/cbarillot" target="_blank">Christian Barillot 1 noforc>1 1 / trong> ="nav nink5a31e14aa6cfe v> js/utile.js?3_20171120"> cryptM <("", nink5a31e14aa6cfe , lyphicon-'hite"> Connenvelope' ")ml { pos ALr: /'_blank'" >Rrget="_bla ScientifV ss" m> rget="_blv> >iv> r: rt_listAPP-O6Y5HZ0000-0002-1589-7696 m> rt_listAPP-O6Y5HZ0000-0002-1589-7696v> >iv> 1 "col-md-12"> uctid">1 k"> uctid">11 d'informations <3ic/Export_liste_publi_HAL" j > ( RISA, campu BeyDiieu F-35042 trlabe labe labe k="$('#fo S r: 11get abe s<"> INSERMmat tiout N" al la r: U746 j > (101, ru Tolbiac, 75013 Ptien - Fr ana) >trlabe labe labe k="$('#fo S r: 303623 abe s<"> I ria j > (Campu beyDiieu 35042 trlabe labe labe k="$('#fo S r: 419153 abe s<"> I ria at tiout N" al j > (D VoluceyD Roclesicourt atBP 105 78153 Le Chesnay Ce - Fr ana) >trlabe labe labe k="$('#fo S r: 30000t abe s<"> RISA-D5matSIGNAUX ET IMAGES NUMÉRIQUES, ROBOTIQUE j > (Fr ana) >trlabe labe labe k="$('#fo S r: 419366 abe s<"> RISA at tiout (Avenu u général Lealizc Campu BeyDiieu 35042 trlabe labe labe k="$('#fo S r: 105128 abe s<"> UBrmatUni (BP 92116mat56321 Lorcts. ce - Fr ana) >trlabe labe labe k="$('#fo S r: 172265 abe s<"> CNRrmatfique dN" al la r: UMR6074 j > (Fr ana) >trlabe labe labe k="$('#fo S r: 4415et abe s<"> ENS j > (Campu Ker Lannd- avenu Robank /chumand- 35170 Bruz - Fr ana) >trlabe labe labe k="$('#fo S r: 247362 abe s<"> I ria at tiout N" al j > (D VoluceyD Roclesicourt atBP 105 78153 Le Chesnay Ce - Fr ana) >trlabe labe labe k="$('#fo S r: 30000t abe s<"> INSA at tiout N" al s A" tiqué sv> r: - (Fr ana) >trlabe labe labe k="$('#fo S r: 301232 abe s<"> Télécom Br v> j > (Tant opôl Br st-Iroise CS 83818 29238 BREST Ce 3 - Fr ana) >trlabe labe labe k="$('#fo S r: 301262 abe s<"> UR1matUni j > (2 ru u ThabormatfS 46510d- 35065 trlabe labe labe k="$('#fo S r: 105160 abe s<"> eS élecv> j > (3, ru Joliot Curie, Platea>trlabe labe labe k="$('#fo S r: 411575 abe s<"> uctid">11inserm-01127692, version 1Détails
k"> uctid">11 trong>A (MS) i / trong>r: with heterogeneous evolution among the patients. Quantitative analysis of longitudinal Magnetic Resonance Images (MRI) provides a spatial analysis of the brain tissues which may lead to the discovery of biomarkers of disease evolution. Better understanding of the disease will lead to a better discovery of pathogenic mechanisms, allowing for patient-adapted therapeutic strategies. To characterize MS lesions, we propose a novel paradigm to detect white matter lesions based on a statistical framework. It aims at studying the benefits of using multi-channel MRI to detect statistically significant differences between each individual MS patient and a database of control subjects. This framework consists in two components. First, intensity standardization is conducted to minimize the inter-subject intensity difference arising from variability of the acquisition process and different scanners. The intensity normalization maps parameters obtained using a robust Gaussian Mixture Model (GMM) estimation not affected by the presence of MS lesions. The second part studies the comparison of multi-channel MRI of MS patients with respect to an atlas built from the control subjects, thereby allowing us to look for differences in normal appearing white matter, in and around the lesions of each patient. Experimental results demonstrate that our technique accurately detects significant differences in lesions consequently improving the results of MS lesion detection." /> 1
uctid">1 k"> uctid">11 trong>Kation ; / trong>r: > 692, version 1clerosis ; Statistsics 1 < href="/search/index/q/*/structId_i/11111lection zation _t/nsity-Normalized M="_blank"> > 692, version 1nsity-Normalized M 1 < href="/search/index/q/*/structId_i/11111lection zation _t/="og:descri="_blank"> > 692, version 1="og:descri 1 < href="/search/index/q/*/structId_i/11111lection zation _t/n a="_blank"> > 692, version 1nRI 1
1 < trong>T > / trong>r: dth="100%" ceabe labe kCOMM labe kputton> Cimg src="/img dividngrèsru
SPIE MepscaNlysison. , Feb blic, Or do,tUnihe s="ogesru
nofo< trong>D trong>r: dal-naquote : 8px;" g" /:65own -15own">w" href="/search/index
/q/*/authIdHal_s/cbarillot" leles0_ _s/sdv m> du Vivtly [q-ter]v> r/ /q/*/authIdHal_s/cbarillot" leles1_ _s/sdv.neu m> Neuzed [q-ter.NC]v> rdiv class="struct"> / trong>
noforch/ u
1 u
inserm-01127692, version 1lyphicon-white"> Connth- Voirructs"> u="nofollow" '.auh _201-ceisintd align= id="glonon"/pu-danger">inserm-01127692, version 1lyphicon-white"> Connth- Mas/scrructs"> u="nofollow" "col-md ascriss u
s _201-ceisinton" class= '.ausmall small btllow" a">inserm-011276nav '.autype7692, version 1="structid">1 Connth- Voir uabtllow" a">inserm-011276nav '.auh 7692, version 1="structid">1 Connth- Mas/scr uabtllow" a">inserm-011276> 692, version 11 Conn"butload Téléch r uabtllow" u="nofollow u="nofollow "col-md ascriss u ontainer"/ Id); tyle> ')) tyle> 111 Id:> Id,g"> Reroving :'vin="nReroving t'}, uctid">1 }tml { posile> iv> 7692">Connexion avec ORCID rbrblicv> criprib : < href="/search/index/q/*/structId_i/111111111111111111111111111111111111111111cbarillot" criprib paId tar471t="_blank"> > <="nav nink5a31e14aa9bc5 v> js/utile.js?3_20171120"> cryptM <("bpbzzbjv@tznvy.pbz", nink5a31e14aa9bc5 , ")ml { pos>
Dernièed modquentc/Exp : mercvert 29er lembed -Ins - 15:41:55rbrblicv> D (s) .targeé(s) : lunrt ns avril -Ins - 03:40:20rbrblicv> >iv> ru
1Fiarg Mv> ruh3>
1 "col-md-12"> lass="col-md 1 aucbarill i>
td> /thumb title="Cent6057530 " "col-mdtd>-thumbn <"hidden" name="identifia class="collapse navbar-collapse" id="nav-servi
den" name="identifia aucbarill i> n den" name="identifia ass=Fiarg Mix docutectplink'(l t) . (s)
noforch/index/q/*/st name="identifia
struct"> struct"> struct">
n -->
t h3iv classwid lass="n I/> nruh3> trong>l u ontainer

n r| aucbarill RISA-D5 m> n r| aucbarillUR1-UFR-ISTIC m> n r| aucbarillCENTRALESUPELEC m> n . SPIE MepscaNlysison. , Feb blic, Or do,tUnihe s="oges. a"> nc/Export_liste_publi_HAL_V3_aouD 〈D 〉v> rdiv u ontainer

>BibTeX uabtllow" aiv classwid le> >TEI uabtllow" aiv classwid le> >DC uabtllow" aiv classwid le> >DChal+s uabtllow" aiv classwid le>EndNo u ontainer
li sdds li="https://www.ccsd.cnrs.fsdds csd.cnrs.fsdds csd.cnrs.fsdds csd.cnrs.fsdds csd.cnrs.fsdds li=" uabtllow" u
struct"> "col-md-12"> "col-md align=ottom b3em" trong> a lter tics trong>l
< labe labe k="$('#fo 305 h2
"col-md align=ottom b3em" trong>Téléch fiarg Mi trong>l
< labe labe k="$('#fo 20t h2
$(s.'.au-equ>m-0112').ors-d(funic/Ex()tyle> tai').h ();yle> });yyle> funic/ExpninkifyElem (selector); tyle> funic/Expails')">Déselector1, selector2); tyle> funic/ExptypeDocMnterame;(docid); tyle> ')) tyle> 111 id:> id}, uctid">1 }tml { posi
csd> /sup brblicv> sd>mailto:bli_sup > td> td>/ellpkc ti-foosin.jpg" positio20 alt="Lllp CCSD });yml { posi .lontc/Ex.proto ) ? r/" onclipiwik-bli_ title="Centf: ort_listpiwik-bli_ title="Cent);yle>
.write(dei=htURI("%3C/ ax-config'le> r/ TS) kin(pkBaseURL + ""iwik.php", 29);yle> LinkTS) king();yle> TS) kin(pkBaseURL + ""iwik.php", 92);yle> LinkTS) king();yle> ax-config td> /" onclipiwik-bli_ title="Cent"iwik.php?idsite=29si align= : alt=" /no-configyle> axbody ruquiv>