Predicting Kidney Graft Survival Using Machine Learning Methods: Prediction Model Development and Feature Significance Analysis Study.
dimensionality reduction
feature sensitivity analysis
kidney transplantation
machine learning
predictive modeling
survival prediction
Journal
Journal of medical Internet research
ISSN: 1438-8871
Titre abrégé: J Med Internet Res
Pays: Canada
ID NLM: 100959882
Informations de publication
Date de publication:
27 08 2021
27 08 2021
Historique:
received:
04
01
2021
accepted:
06
05
2021
revised:
10
03
2021
entrez:
27
8
2021
pubmed:
28
8
2021
medline:
27
10
2021
Statut:
epublish
Résumé
Kidney transplantation is the optimal treatment for patients with end-stage renal disease. Short- and long-term kidney graft survival is influenced by a number of donor and recipient factors. Predicting the success of kidney transplantation is important for optimizing kidney allocation. The aim of this study was to predict the risk of kidney graft failure across three temporal cohorts (within 1 year, within 5 years, and after 5 years following a transplant) based on donor and recipient characteristics. We analyzed a large data set comprising over 50,000 kidney transplants covering an approximate 20-year period. We applied machine learning-based classification algorithms to develop prediction models for the risk of graft failure for three different temporal cohorts. Deep learning-based autoencoders were applied for data dimensionality reduction, which improved the prediction performance. The influence of features on graft survival for each cohort was studied by investigating a new nonoverlapping patient stratification approach. Our models predicted graft survival with area under the curve scores of 82% within 1 year, 69% within 5 years, and 81% within 17 years. The feature importance analysis elucidated the varying influence of clinical features on graft survival across the three different temporal cohorts. In this study, we applied machine learning to develop risk prediction models for graft failure that demonstrated a high level of prediction performance. Acknowledging that these models performed better than those reported in the literature for existing risk prediction tools, future studies will focus on how best to incorporate these prediction models into clinical care algorithms to optimize the long-term health of kidney recipients.
Sections du résumé
BACKGROUND
Kidney transplantation is the optimal treatment for patients with end-stage renal disease. Short- and long-term kidney graft survival is influenced by a number of donor and recipient factors. Predicting the success of kidney transplantation is important for optimizing kidney allocation.
OBJECTIVE
The aim of this study was to predict the risk of kidney graft failure across three temporal cohorts (within 1 year, within 5 years, and after 5 years following a transplant) based on donor and recipient characteristics. We analyzed a large data set comprising over 50,000 kidney transplants covering an approximate 20-year period.
METHODS
We applied machine learning-based classification algorithms to develop prediction models for the risk of graft failure for three different temporal cohorts. Deep learning-based autoencoders were applied for data dimensionality reduction, which improved the prediction performance. The influence of features on graft survival for each cohort was studied by investigating a new nonoverlapping patient stratification approach.
RESULTS
Our models predicted graft survival with area under the curve scores of 82% within 1 year, 69% within 5 years, and 81% within 17 years. The feature importance analysis elucidated the varying influence of clinical features on graft survival across the three different temporal cohorts.
CONCLUSIONS
In this study, we applied machine learning to develop risk prediction models for graft failure that demonstrated a high level of prediction performance. Acknowledging that these models performed better than those reported in the literature for existing risk prediction tools, future studies will focus on how best to incorporate these prediction models into clinical care algorithms to optimize the long-term health of kidney recipients.
Identifiants
pubmed: 34448704
pii: v23i8e26843
doi: 10.2196/26843
pmc: PMC8433864
doi:
Types de publication
Journal Article
Langues
eng
Sous-ensembles de citation
IM
Pagination
e26843Informations de copyright
©Syed Asil Ali Naqvi, Karthik Tennankore, Amanda Vinson, Patrice C Roy, Syed Sibte Raza Abidi. Originally published in the Journal of Medical Internet Research (https://www.jmir.org), 27.08.2021.
Références
Nephrology (Carlton). 2019 Jun;24(6):591-598
pubmed: 30536674
J Nucl Cardiol. 2014 Aug;21(4):686-94
pubmed: 24810431
BMJ Clin Evid. 2007 Oct 17;2007:
pubmed: 19450356
Int J Med Inform. 2019 Oct;130:103957
pubmed: 31472443
N Engl J Med. 2004 Sep 23;351(13):1296-305
pubmed: 15385656
Kidney Int. 2016 Mar;89(3):648-58
pubmed: 26880458
J Urol. 2009 Mar;181(3):1248-55
pubmed: 19167732
Ann Thorac Surg. 2011 Aug;92(2):520-7; discussion 527
pubmed: 21683337
Clin J Am Soc Nephrol. 2014 Jul;9(7):1153-5
pubmed: 24903388
Mol Pharm. 2016 May 2;13(5):1445-54
pubmed: 27007977
Transplant Direct. 2018 Dec 19;5(1):e414
pubmed: 30656212
Int J Med Inform. 2018 Nov;119:125-133
pubmed: 30342680
Artif Intell Med. 2017 Mar;77:1-11
pubmed: 28545607
Annu Int Conf IEEE Eng Med Biol Soc. 2017 Jul;2017:74-77
pubmed: 29059814
Can J Kidney Health Dis. 2018 Jul 27;5:2054358118791148
pubmed: 30083367
Transplantation. 2009 Jul 27;88(2):231-6
pubmed: 19623019
Annu Int Conf IEEE Eng Med Biol Soc. 2016 Aug;2016:3290-3293
pubmed: 28269008
Am J Nephrol. 2012;36(6):561-9
pubmed: 23221105
Nat Rev Nephrol. 2012 Mar 27;8(6):325-31
pubmed: 22450438
PLoS One. 2018 Mar 28;13(3):e0194985
pubmed: 29590219
Am J Transplant. 2018 May;18(5):1168-1176
pubmed: 29451350
J Biomed Inform. 2008 Dec;41(6):944-52
pubmed: 18442951
Transplant Proc. 2010 May;42(4):1130-3
pubmed: 20534242
BMJ. 2019 Sep 17;366:l4923
pubmed: 31530561
Sci Rep. 2017 Aug 21;7(1):8904
pubmed: 28827646
Artif Intell Med. 2011 Oct;53(2):107-18
pubmed: 21821401