A large, curated, open-source stroke neuroimaging dataset to improve lesion segmentation algorithms - Inria - Institut national de recherche en sciences et technologies du numérique Accéder directement au contenu
Article Dans Une Revue Scientific Data Année : 2022

A large, curated, open-source stroke neuroimaging dataset to improve lesion segmentation algorithms

1 USC - University of Southern California
2 PITT - University of Pittsburgh
3 Clinical and Behavioral Neurology - Neuroscienze e riabilitazione, IRCCS Fondazione Santa Lucia [Roma]
4 Emory University School of Medicine
5 University of British Columbia [Vancouver]
6 University of Melbourne
7 Child Mind Institute
8 Department Biostatistics University of North Carolina
9 USP - Universidade de São Paulo = University of São Paulo
10 UC - University of California
11 University of Barcelona
12 TUM - Technische Universität Munchen - Technical University Munich - Université Technique de Munich
13 Universität Greifswald - University of Greifswald
14 University of Texas at Austin [Austin]
15 Duke University [Durham]
16 NKI - Nathan S. Kline Institute for Psychiatric Research
17 NYU Grossman School of Medicine - New York University School of Medicine
18 Imperial College London
19 MIND - Modèles et inférence pour les données de Neuroimagerie
20 MUSC - Medical University of South Carolina [Charleston]
21 Wake Forest School of Medicine [Winston-Salem]
22 University of South Australia [Adelaide]
23 The Florey Institute of Neuroscience and Mental Health
24 TJU - Tianjin University
25 University of Toronto
26 UB - Universitat de Barcelona
27 OsloMet - Oslo Metropolitan University
28 University of Michigan [Ann Arbor]
29 Emory University [Atlanta, GA]
30 UNIBE - Universität Bern / University of Bern
31 University of Waterloo [Waterloo]
32 Charité - UniversitätsMedizin = Charité - University Hospital [Berlin]
33 Oslo University Hospital [Oslo]
Sook-Lei Liew
Giuseppe Barisano
Tyler Ard
Lara Boyd
Cathrin Buetefisch
Lei Cao
  • Fonction : Auteur
Jingchun Liu
  • Fonction : Auteur
Feroze Mohamed
  • Fonction : Auteur
Amy Pienta
Andrew Robertson
Gregory Thielman
  • Fonction : Auteur
Lars Westlye
George Wittenberg
Chunshui Yu
  • Fonction : Auteur

Résumé

Accurate lesion segmentation is critical in stroke rehabilitation research for the quantification of lesion burden and accurate image processing. Current automated lesion segmentation methods for T1-weighted (T1w) MRIs, commonly used in stroke research, lack accuracy and reliability. Manual segmentation remains the gold standard, but it is time-consuming, subjective, and requires neuroanatomical expertise. We previously released an open-source dataset of stroke T1w MRIs and manually-segmented lesion masks (ATLAS v1.2, N = 304) to encourage the development of better algorithms. However, many methods developed with ATLAS v1.2 report low accuracy, are not publicly accessible or are improperly validated, limiting their utility to the field. Here we present ATLAS v2.0 (N = 1271), a larger dataset of T1w MRIs and manually segmented lesion masks that includes training (n = 655), test (hidden masks, n = 300), and generalizability (hidden MRIs and masks, n = 316) datasets. Algorithm development using this larger sample should lead to more robust solutions; the hidden datasets allow for unbiased performance evaluation via segmentation challenges. We anticipate that ATLAS v2.0 will lead to improved algorithms, facilitating large-scale stroke research.

Dates et versions

hal-03808195 , version 1 (10-10-2022)

Identifiants

Citer

Sook-Lei Liew, Bethany Lo, Miranda Donnelly, Artemis Zavaliangos-Petropulu, Jessica Jeong, et al.. A large, curated, open-source stroke neuroimaging dataset to improve lesion segmentation algorithms. Scientific Data , 2022, 9 (1), pp.320. ⟨10.1038/s41597-022-01401-7⟩. ⟨hal-03808195⟩
28 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More