Detection of cerebral aneurysms using artificial intelligence: a systematic review and meta-analysis.
Aneurysm
Angiography
Brain
CT Angiography
Magnetic Resonance Angiography
artificial intelligence
deep learning
machine learning
Journal
Journal of neurointerventional surgery
ISSN: 1759-8486
Titre abrégé: J Neurointerv Surg
Pays: England
ID NLM: 101517079
Informations de publication
Date de publication:
Mar 2023
Mar 2023
Historique:
received:
29
07
2022
accepted:
11
10
2022
pubmed:
15
11
2022
medline:
17
2
2023
entrez:
14
11
2022
Statut:
ppublish
Résumé
Subarachnoid hemorrhage from cerebral aneurysm rupture is a major cause of morbidity and mortality. Early aneurysm identification, aided by automated systems, may improve patient outcomes. Therefore, a systematic review and meta-analysis of the diagnostic accuracy of artificial intelligence (AI) algorithms in detecting cerebral aneurysms using CT, MRI or DSA was performed. MEDLINE, Embase, Cochrane Library and Web of Science were searched until August 2021. Eligibility criteria included studies using fully automated algorithms to detect cerebral aneurysms using MRI, CT or DSA. Following Preferred Reporting Items for Systematic Reviews and Meta-Analysis: Diagnostic Test Accuracy (PRISMA-DTA), articles were assessed using Quality Assessment of Diagnostic Accuracy Studies 2 (QUADAS-2). Meta-analysis included a bivariate random-effect model to determine pooled sensitivity, specificity, and area under the receiver operator characteristic curve (ROC-AUC). CRD42021278454. 43 studies were included, and 41/43 (95%) were retrospective. 34/43 (79%) used AI as a standalone tool, while 9/43 (21%) used AI assisting a reader. 23/43 (53%) used deep learning. Most studies had high bias risk and applicability concerns, limiting conclusions. Six studies in the standalone AI meta-analysis gave (pooled) 91.2% (95% CI 82.2% to 95.8%) sensitivity; 16.5% (95% CI 9.4% to 27.1%) false-positive rate (1-specificity); 0.936 ROC-AUC. Five reader-assistive AI studies gave (pooled) 90.3% (95% CI 88.0% - 92.2%) sensitivity; 7.9% (95% CI 3.5% to 16.8%) false-positive rate; 0.910 ROC-AUC. AI has the potential to support clinicians in detecting cerebral aneurysms. Interpretation is limited due to high risk of bias and poor generalizability. Multicenter, prospective studies are required to assess AI in clinical practice.
Sections du résumé
BACKGROUND
BACKGROUND
Subarachnoid hemorrhage from cerebral aneurysm rupture is a major cause of morbidity and mortality. Early aneurysm identification, aided by automated systems, may improve patient outcomes. Therefore, a systematic review and meta-analysis of the diagnostic accuracy of artificial intelligence (AI) algorithms in detecting cerebral aneurysms using CT, MRI or DSA was performed.
METHODS
METHODS
MEDLINE, Embase, Cochrane Library and Web of Science were searched until August 2021. Eligibility criteria included studies using fully automated algorithms to detect cerebral aneurysms using MRI, CT or DSA. Following Preferred Reporting Items for Systematic Reviews and Meta-Analysis: Diagnostic Test Accuracy (PRISMA-DTA), articles were assessed using Quality Assessment of Diagnostic Accuracy Studies 2 (QUADAS-2). Meta-analysis included a bivariate random-effect model to determine pooled sensitivity, specificity, and area under the receiver operator characteristic curve (ROC-AUC).
PROSPERO
UNASSIGNED
CRD42021278454.
RESULTS
RESULTS
43 studies were included, and 41/43 (95%) were retrospective. 34/43 (79%) used AI as a standalone tool, while 9/43 (21%) used AI assisting a reader. 23/43 (53%) used deep learning. Most studies had high bias risk and applicability concerns, limiting conclusions. Six studies in the standalone AI meta-analysis gave (pooled) 91.2% (95% CI 82.2% to 95.8%) sensitivity; 16.5% (95% CI 9.4% to 27.1%) false-positive rate (1-specificity); 0.936 ROC-AUC. Five reader-assistive AI studies gave (pooled) 90.3% (95% CI 88.0% - 92.2%) sensitivity; 7.9% (95% CI 3.5% to 16.8%) false-positive rate; 0.910 ROC-AUC.
CONCLUSION
CONCLUSIONS
AI has the potential to support clinicians in detecting cerebral aneurysms. Interpretation is limited due to high risk of bias and poor generalizability. Multicenter, prospective studies are required to assess AI in clinical practice.
Identifiants
pubmed: 36375834
pii: jnis-2022-019456
doi: 10.1136/jnis-2022-019456
pmc: PMC9985742
doi:
Types de publication
Meta-Analysis
Systematic Review
Journal Article
Langues
eng
Sous-ensembles de citation
IM
Pagination
262-271Subventions
Organisme : Medical Research Council
ID : MR/W021684/1
Pays : United Kingdom
Informations de copyright
© Author(s) (or their employer(s)) 2023. Re-use permitted under CC BY. Published by BMJ.
Déclaration de conflit d'intérêts
Competing interests: Microvention Core laboratory for neurointerventional study. Payment to TCB. Not related to AI. Siemens Healthineers Speakers Bureau Educational lecture on brain tumours. Payment to TCB. Not related to aneurysms. Medtronic Speakers Bureau Educational lecture on an independent non-industry sponsored study published in AJNR on Pipeline Flow Diverter device. Payment to TCB. Not related to AI.
Références
Biomed Tech (Berl). 2018 Jul 26;63(4):445-452
pubmed: 28672767
Radiology. 2019 Jan;290(1):187-194
pubmed: 30351253
Biomed Eng Online. 2020 May 29;19(1):38
pubmed: 32471439
JAMA Netw Open. 2019 Jun 5;2(6):e195600
pubmed: 31173130
J Clin Epidemiol. 2005 Oct;58(10):982-90
pubmed: 16168343
Int J Comput Assist Radiol Surg. 2019 Aug;14(8):1353-1364
pubmed: 31119487
Med Phys. 2016 Feb;43(2):951-60
pubmed: 26843255
Jpn J Radiol. 2021 Jul;39(7):652-658
pubmed: 33638771
AJNR Am J Neuroradiol. 2019 Jan;40(1):25-32
pubmed: 30573461
Neuroradiol J. 2020 Aug;33(4):311-317
pubmed: 32633602
J Digit Imaging. 2019 Oct;32(5):808-815
pubmed: 30511281
AJR Am J Roentgenol. 2013 Sep;201(3):611-7
pubmed: 23971454
Lancet. 2007 Mar 17;369(9565):904
pubmed: 17368147
Clin Radiol. 2020 Jan;75(1):20-32
pubmed: 31371027
Int J Comput Assist Radiol Surg. 2020 Apr;15(4):715-723
pubmed: 32056126
Biomed Eng Online. 2019 Nov 14;18(1):110
pubmed: 31727057
Neuroradiology. 2021 Dec;63(12):1985-1994
pubmed: 33837806
JAMA. 2018 Jan 23;319(4):388-396
pubmed: 29362800
Neurol Clin Pract. 2013 Apr;3(2):99-108
pubmed: 23914319
AJNR Am J Neuroradiol. 2014 Oct;35(10):1897-902
pubmed: 24924543
Radiology. 2021 Jan;298(1):164-165
pubmed: 33146581
IEEE Trans Med Imaging. 2020 May;39(5):1448-1458
pubmed: 31689186
Lancet Neurol. 2011 Jul;10(7):626-36
pubmed: 21641282
Med Image Anal. 2010 Apr;14(2):149-59
pubmed: 20004607
Br J Radiol. 2020 Feb 1;93(1106):20190855
pubmed: 31965813
Emerg Med J. 2019 Jul;36(7):431-434
pubmed: 31302605
J Digit Imaging. 2011 Feb;24(1):86-95
pubmed: 19937083
Ann Intern Med. 2011 Oct 18;155(8):529-36
pubmed: 22007046
Biomed Eng Online. 2020 Sep 15;19(1):73
pubmed: 32933534
J Clin Epidemiol. 2020 Feb;118:42-54
pubmed: 31698064
Eur J Radiol. 2022 Oct;155:110457
pubmed: 35961157
Front Oncol. 2022 Jan 31;12:799662
pubmed: 35174084
Clin Kidney J. 2020 Nov 24;14(1):49-58
pubmed: 33564405
Res Synth Methods. 2010 Apr;1(2):97-111
pubmed: 26061376
AJNR Am J Neuroradiol. 2016 Jun;37(6):1038-43
pubmed: 26892988
Interv Neuroradiol. 2021 Oct;27(5):648-657
pubmed: 33715500
Acad Radiol. 2015 Sep;22(9):1191-8
pubmed: 26210525
Interv Neuroradiol. 2022 May 13;:15910199221097475
pubmed: 35549574
Radiol Artif Intell. 2020 Mar 25;2(2):e200029
pubmed: 33937821
AJNR Am J Neuroradiol. 2021 Oct;42(10):1769-1775
pubmed: 34385143
Lancet Neurol. 2009 Jul;8(7):635-42
pubmed: 19501022
Acad Radiol. 2004 Oct;11(10):1093-104
pubmed: 15530802
J Digit Imaging. 2017 Oct;30(5):629-639
pubmed: 28405834
Int J Comput Assist Radiol Surg. 2019 Dec;14(12):2095-2107
pubmed: 30859456
Eur Radiol. 2020 Nov;30(11):5785-5793
pubmed: 32474633
Med Phys. 2014 Sep;41(9):091904
pubmed: 25186391
J Neurointerv Surg. 2020 Oct;12(10):1023-1027
pubmed: 32471827
Comput Med Imaging Graph. 2021 Apr;89:101888
pubmed: 33690001
Stat Med. 2008 Feb 28;27(5):687-97
pubmed: 17611957
Med Phys. 2006 Feb;33(2):394-401
pubmed: 16532946
AJNR Am J Neuroradiol. 2020 Mar;41(3):373-379
pubmed: 32165361
Jpn J Radiol. 2021 Nov;39(11):1039-1048
pubmed: 34125368
Radiology. 2018 Mar;286(3):800-809
pubmed: 29309734
Int J Technol Assess Health Care. 2012 Apr;28(2):138-44
pubmed: 22559755
Sci Rep. 2020 Dec 11;10(1):21799
pubmed: 33311535
Clin Chem. 2017 Aug;63(8):1331-1341
pubmed: 28630237
Acad Radiol. 2022 Mar;29 Suppl 3:S201-S214
pubmed: 34376335
Radiology. 2021 Jan;298(1):155-163
pubmed: 33141003
Radiology. 2005 Nov;237(2):605-10
pubmed: 16179404
Eur Radiol. 2022 Jan;32(1):725-736
pubmed: 34286375
AJR Am J Roentgenol. 2008 Feb;190(2):459-65
pubmed: 18212233
Neuroimage. 2021 Sep;238:118216
pubmed: 34052465
Health Care Manage Rev. 2011 Oct-Dec;36(4):338-48
pubmed: 21685794
Clin Neuroradiol. 2020 Sep;30(3):591-598
pubmed: 31227844
J Magn Reson Imaging. 2018 Apr;47(4):948-953
pubmed: 28836310
Nat Commun. 2020 Nov 30;11(1):6090
pubmed: 33257700