Machine learning-based prediction of clinical outcomes after first-ever ischemic stroke.
cognitive impairment
functional outcome
machine learning
mortality
outcome prediction
post-stroke depression
stroke
Journal
Frontiers in neurology
ISSN: 1664-2295
Titre abrégé: Front Neurol
Pays: Switzerland
ID NLM: 101546899
Informations de publication
Date de publication:
2023
2023
Historique:
received:
02
12
2022
accepted:
31
01
2023
entrez:
10
3
2023
pubmed:
11
3
2023
medline:
11
3
2023
Statut:
epublish
Résumé
Accurate prediction of clinical outcomes in individual patients following acute stroke is vital for healthcare providers to optimize treatment strategies and plan further patient care. Here, we use advanced machine learning (ML) techniques to systematically compare the prediction of functional recovery, cognitive function, depression, and mortality of first-ever ischemic stroke patients and to identify the leading prognostic factors. We predicted clinical outcomes for 307 patients (151 females, 156 males; 68 ± 14 years) from the PROSpective Cohort with Incident Stroke Berlin study using 43 baseline features. Outcomes included modified Rankin Scale (mRS), Barthel Index (BI), Mini-Mental State Examination (MMSE), Modified Telephone Interview for Cognitive Status (TICS-M), Center for Epidemiologic Studies Depression Scale (CES-D) and survival. The ML models included a Support Vector Machine with a linear kernel and a radial basis function kernel as well as a Gradient Boosting Classifier based on repeated 5-fold nested cross-validation. The leading prognostic features were identified using Shapley additive explanations. The ML models achieved significant prediction performance for mRS at patient discharge and after 1 year, BI and MMSE at patient discharge, TICS-M after 1 and 3 years and CES-D after 1 year. Additionally, we showed that National Institutes of Health Stroke Scale (NIHSS) was the top predictor for most functional recovery outcomes as well as education for cognitive function and depression. Our machine learning analysis successfully demonstrated the ability to predict clinical outcomes after first-ever ischemic stroke and identified the leading prognostic factors that contribute to this prediction.
Sections du résumé
Background
UNASSIGNED
Accurate prediction of clinical outcomes in individual patients following acute stroke is vital for healthcare providers to optimize treatment strategies and plan further patient care. Here, we use advanced machine learning (ML) techniques to systematically compare the prediction of functional recovery, cognitive function, depression, and mortality of first-ever ischemic stroke patients and to identify the leading prognostic factors.
Methods
UNASSIGNED
We predicted clinical outcomes for 307 patients (151 females, 156 males; 68 ± 14 years) from the PROSpective Cohort with Incident Stroke Berlin study using 43 baseline features. Outcomes included modified Rankin Scale (mRS), Barthel Index (BI), Mini-Mental State Examination (MMSE), Modified Telephone Interview for Cognitive Status (TICS-M), Center for Epidemiologic Studies Depression Scale (CES-D) and survival. The ML models included a Support Vector Machine with a linear kernel and a radial basis function kernel as well as a Gradient Boosting Classifier based on repeated 5-fold nested cross-validation. The leading prognostic features were identified using Shapley additive explanations.
Results
UNASSIGNED
The ML models achieved significant prediction performance for mRS at patient discharge and after 1 year, BI and MMSE at patient discharge, TICS-M after 1 and 3 years and CES-D after 1 year. Additionally, we showed that National Institutes of Health Stroke Scale (NIHSS) was the top predictor for most functional recovery outcomes as well as education for cognitive function and depression.
Conclusion
UNASSIGNED
Our machine learning analysis successfully demonstrated the ability to predict clinical outcomes after first-ever ischemic stroke and identified the leading prognostic factors that contribute to this prediction.
Identifiants
pubmed: 36895902
doi: 10.3389/fneur.2023.1114360
pmc: PMC9990416
doi:
Types de publication
Journal Article
Langues
eng
Pagination
1114360Informations de copyright
Copyright © 2023 Fast, Temuulen, Villringer, Kufner, Ali, Siebert, Huo, Piper, Sperber, Liman, Endres and Ritter.
Déclaration de conflit d'intérêts
ME reports grants from Bayer and fees paid to the Charité from Abbot, Amgen, AstraZeneca, Bayer, 296 Boehringer Ingelheim, BMS, Daiishi Sankyo, Sanofi, Novartis, Pfizer, all outside the submitted work. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Références
PLoS One. 2018 Dec 20;13(12):e0208594
pubmed: 30571691
JAMA. 1993 May 12;269(18):2386-91
pubmed: 8479064
Psychosomatics. 2009 Nov-Dec;50(6):563-9
pubmed: 19996226
Stroke. 2019 May;50(5):1263-1265
pubmed: 30890116
Front Neurol. 2018 May 07;9:308
pubmed: 29867722
Obes Res Clin Pract. 2011 Oct-Dec;5(4):e267-360
pubmed: 24331129
Int J Methods Psychiatr Res. 2011 Mar;20(1):40-9
pubmed: 21499542
BMC Proc. 2011 May 27;5 Suppl 3:S11
pubmed: 21624167
J Alzheimers Dis. 2020;75(3):717-728
pubmed: 32333585
J Neurol. 2009 Dec;256(12):2003-8
pubmed: 19609738
IEEE/ACM Trans Comput Biol Bioinform. 2018 Nov-Dec;15(6):1953-1959
pubmed: 29994736
Sci Rep. 2021 Jul 8;11(1):14168
pubmed: 34239011
Lancet Neurol. 2019 May;18(5):417-418
pubmed: 30871943
IEEE Rev Biomed Eng. 2021;14:156-180
pubmed: 32746371
J Alzheimers Dis. 2015;49(4):1169-77
pubmed: 26599056
Int J Stroke. 2014 Dec;9(8):1017-25
pubmed: 25117911
Circulation. 2011 Feb 22;123(7):739-49
pubmed: 21300951
Stroke. 2010 May;41(5):992-5
pubmed: 20360551
Sci Rep. 2020 Nov 11;10(1):19571
pubmed: 33177575
Arch Intern Med. 2012 Nov 12;172(20):1548-56
pubmed: 23147454
NPJ Digit Med. 2020 Feb 6;3:17
pubmed: 32047862
Ann Transl Med. 2019 Apr;7(7):152
pubmed: 31157273
Neurology. 2012 Dec 4;79(23):2293; author reply 2293-4
pubmed: 23212191
Dement Geriatr Cogn Disord. 2008;26(4):356-63
pubmed: 18852488
NeuroRehabilitation. 2012;30(4):315-22
pubmed: 22672946
Int J Stroke. 2013 Aug;8(6):484-90
pubmed: 22928669
PLoS One. 2013;8(3):e58067
pubmed: 23469262
Stroke. 2001 Jun;32(6):1318-22
pubmed: 11387493
Cochrane Database Syst Rev. 2016 Sep 01;9:CD002293
pubmed: 27581994
Front Neurol. 2020 Nov 19;11:539509
pubmed: 33329298
Appl Nurs Res. 2000 Nov;13(4):209-13
pubmed: 11078787
J Geriatr Psychiatry Neurol. 2009 Jun;22(2):103-9
pubmed: 19417219
Elife. 2022 May 26;11:
pubmed: 35616520
PLoS One. 2020 Jun 12;15(6):e0234722
pubmed: 32530947
J Clin Lab Anal. 2020 Jul;34(7):e23301
pubmed: 32196744
Sci Rep. 2017 Sep 29;7(1):12441
pubmed: 28963553
Pathophysiology. 2010 Jun;17(3):197-218
pubmed: 20074922
Clin J Am Soc Nephrol. 2012 Aug;7(8):1355-64
pubmed: 22679181
Nat Biomed Eng. 2018 Oct;2(10):749-760
pubmed: 31001455
Diagnostics (Basel). 2021 Jan 06;11(1):
pubmed: 33419013
J Cheminform. 2014 Mar 29;6(1):10
pubmed: 24678909
Nat Med. 2019 Jan;25(1):44-56
pubmed: 30617339
Hum Brain Mapp. 2002 Jan;15(1):1-25
pubmed: 11747097
PLoS One. 2014 Feb 04;9(2):e87987
pubmed: 24505342
J Ayub Med Coll Abbottabad. 2021 Apr-Jun;33(2):274-278
pubmed: 34137544
Behav Neurol. 2000;12(4):191-200
pubmed: 11568431
Stroke Res Treat. 2018 Nov 27;2018:3238165
pubmed: 30598741
Front Aging Neurosci. 2017 Jul 11;9:218
pubmed: 28744213
Sci Rep. 2017 May 31;7(1):2507
pubmed: 28566757
NPJ Digit Med. 2018 May 8;1:18
pubmed: 31304302
Surg Neurol Int. 2017 Jul 18;8:149
pubmed: 28791192
Br J Nurs. 1992 Jul 23-Aug 13;1(7):325-9
pubmed: 1504565
J Gerontol B Psychol Sci Soc Sci. 2020 Mar 9;75(4):733-742
pubmed: 29718387