Automatic segmentation of the spinal cord nerve rootlets

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F61989592%3A15110%2F24%3A73626512" target="_blank" >RIV/61989592:15110/24:73626512 - isvavai.cz</a>

Výsledek na webu

<a href="https://direct.mit.edu/imag/article/doi/10.1162/imag_a_00218/122601/Automatic-segmentation-of-the-spinal-cord-nerve" target="_blank" >https://direct.mit.edu/imag/article/doi/10.1162/imag_a_00218/122601/Automatic-segmentation-of-the-spinal-cord-nerve</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1162/imag_a_00218" target="_blank" >10.1162/imag_a_00218</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Automatic segmentation of the spinal cord nerve rootlets

Popis výsledku v původním jazyce

Precise identification of spinal nerve rootlets is relevant to delineate spinal levels for the study of functional activity in the spinal cord. The goal of this study was to develop an automatic method for the semantic segmentation of spinal nerve rootlets from T2-weighted magnetic resonance imaging (MRI) scans. Images from two open-access 3T MRIdatasets were used to train a 3D multi-class convolutional neural network using an active learning approach to segment C2-C8 dorsal nerve rootlets. Each output class corresponds o a spinal level. The method was tested on 3T T2-weighted images from three datasets unseen during training to assess inter-site, inter-session, and inter-resolutionvariability. The test Dice score was 0.67 ± 0.16 (mean ± standard deviation across testing images and rootlets levels), suggesting a good performance. The method also demonstrated low inter-vendor and inter-site variability (coefficient of variation ≤ 1.41%), as well as low inter-session variability (coefficient of variation ≤ 1.30%), indicating stable predictions across different MRI vendors, sites, and sessions. The proposed methodology is open-source and readily available in the Spinal Cord Toolbox (SCT) v6.2 and higher.

Název v anglickém jazyce

Automatic segmentation of the spinal cord nerve rootlets

Popis výsledku anglicky

Precise identification of spinal nerve rootlets is relevant to delineate spinal levels for the study of functional activity in the spinal cord. The goal of this study was to develop an automatic method for the semantic segmentation of spinal nerve rootlets from T2-weighted magnetic resonance imaging (MRI) scans. Images from two open-access 3T MRIdatasets were used to train a 3D multi-class convolutional neural network using an active learning approach to segment C2-C8 dorsal nerve rootlets. Each output class corresponds o a spinal level. The method was tested on 3T T2-weighted images from three datasets unseen during training to assess inter-site, inter-session, and inter-resolutionvariability. The test Dice score was 0.67 ± 0.16 (mean ± standard deviation across testing images and rootlets levels), suggesting a good performance. The method also demonstrated low inter-vendor and inter-site variability (coefficient of variation ≤ 1.41%), as well as low inter-session variability (coefficient of variation ≤ 1.30%), indicating stable predictions across different MRI vendors, sites, and sessions. The proposed methodology is open-source and readily available in the Spinal Cord Toolbox (SCT) v6.2 and higher.

Druh

J_ost - Ostatní články v recenzovaných periodicích

CEP obor

—

OECD FORD obor

30210 - Clinical neurology

Projekt

<a href="/cs/project/NU22-04-00024" target="_blank" >NU22-04-00024: Kvantitativní zobrazovací MR parametry jako prediktory průběhu nemyelopatické degenerativní komprese krční míchy: longitudinální studie</a><br>

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Rok uplatnění

2024

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název periodika

Imaging Neuroscience

ISSN

2837-6056

e-ISSN

2837-6056

Svazek periodika

2 (2024)

Číslo periodika v rámci svazku

July 2024

Stát vydavatele periodika

US - Spojené státy americké

Počet stran výsledku

14

Strana od-do

"Article:2024-07"-"07-02"

Kód UT WoS článku

001530558400003

EID výsledku v databázi Scopus

2-s2.0-105009931726