Artificial intelligence for the prevention and prediction of colorectal neoplasms.
Artificial intelligence
Blood data
Colorectal cancer
Machine learning
Screening
Journal
Journal of translational medicine
ISSN: 1479-5876
Titre abrégé: J Transl Med
Pays: England
ID NLM: 101190741
Informations de publication
Date de publication:
03 07 2023
03 07 2023
Historique:
received:
26
03
2023
accepted:
09
06
2023
medline:
5
7
2023
pubmed:
4
7
2023
entrez:
3
7
2023
Statut:
epublish
Résumé
Colonoscopy is a useful as a cancer screening test. However, in countries with limited medical resources, there are restrictions on the widespread use of endoscopy. Non-invasive screening methods to determine whether a patient requires a colonoscopy are thus desired. Here, we investigated whether artificial intelligence (AI) can predict colorectal neoplasia. We used data from physical exams and blood analyses to determine the incidence of colorectal polyp. However, these features exhibit highly overlapping classes. The use of a kernel density estimator (KDE)-based transformation improved the separability of both classes. Along with an adequate polyp size threshold, the optimal machine learning (ML) models' performance provided 0.37 and 0.39 Matthews correlation coefficient (MCC) for the datasets of men and women, respectively. The models exhibit a higher discrimination than fecal occult blood test with 0.047 and 0.074 MCC for men and women, respectively. The ML model can be chosen according to the desired polyp size discrimination threshold, may suggest further colorectal screening, and possible adenoma size. The KDE feature transformation could serve to score each biomarker and background factors (health lifestyles) to suggest measures to be taken against colorectal adenoma growth. All the information that the AI model provides can lower the workload for healthcare providers and be implemented in health care systems with scarce resources. Furthermore, risk stratification may help us to optimize the efficiency of resources for screening colonoscopy.
Sections du résumé
BACKGROUND
Colonoscopy is a useful as a cancer screening test. However, in countries with limited medical resources, there are restrictions on the widespread use of endoscopy. Non-invasive screening methods to determine whether a patient requires a colonoscopy are thus desired. Here, we investigated whether artificial intelligence (AI) can predict colorectal neoplasia.
METHODS
We used data from physical exams and blood analyses to determine the incidence of colorectal polyp. However, these features exhibit highly overlapping classes. The use of a kernel density estimator (KDE)-based transformation improved the separability of both classes.
RESULTS
Along with an adequate polyp size threshold, the optimal machine learning (ML) models' performance provided 0.37 and 0.39 Matthews correlation coefficient (MCC) for the datasets of men and women, respectively. The models exhibit a higher discrimination than fecal occult blood test with 0.047 and 0.074 MCC for men and women, respectively.
CONCLUSION
The ML model can be chosen according to the desired polyp size discrimination threshold, may suggest further colorectal screening, and possible adenoma size. The KDE feature transformation could serve to score each biomarker and background factors (health lifestyles) to suggest measures to be taken against colorectal adenoma growth. All the information that the AI model provides can lower the workload for healthcare providers and be implemented in health care systems with scarce resources. Furthermore, risk stratification may help us to optimize the efficiency of resources for screening colonoscopy.
Identifiants
pubmed: 37400891
doi: 10.1186/s12967-023-04258-5
pii: 10.1186/s12967-023-04258-5
pmc: PMC10318774
doi:
Types de publication
Journal Article
Langues
eng
Sous-ensembles de citation
IM
Pagination
431Informations de copyright
© 2023. The Author(s).
Références
Clin Cancer Res. 2019 May 15;25(10):3006-3015
pubmed: 30979733
Nat Mach Intell. 2020 Jan;2(1):56-67
pubmed: 32607472
Mol Carcinog. 2016 Aug;55(8):1251-61
pubmed: 26333108
Biomed Res Int. 2013;2013:303982
pubmed: 24307992
Cancer Causes Control. 2015 Mar;26(3):467-73
pubmed: 25618792
N Engl J Med. 1993 May 13;328(19):1365-71
pubmed: 8474513
Lancet Oncol. 2005 Nov;6(11):871-6
pubmed: 16257795
N Engl J Med. 2014 Apr 3;370(14):1287-97
pubmed: 24645800
J Gastroenterol Hepatol. 2010 Mar;25(3):562-7
pubmed: 20074156
N Engl J Med. 2012 Feb 23;366(8):687-96
pubmed: 22356322
Gut. 2014 Jul;63(7):1112-9
pubmed: 24385598
Comput Math Methods Med. 2016;2016:6584725
pubmed: 27847543
Biochem Biophys Res Commun. 2014 May 16;447(4):707-14
pubmed: 24769205
CA Cancer J Clin. 2008 May-Jun;58(3):130-60
pubmed: 18322143
Aging (Albany NY). 2016 May;8(5):1021-33
pubmed: 27191382
Cancer Causes Control. 2013 Jun;24(6):1207-22
pubmed: 23563998
Lancet Oncol. 2014 Apr;15(5):489-538
pubmed: 24731404
Lab Med. 2015 Spring;46(2):123-35
pubmed: 25918191
Med Phys. 2008 Apr;35(4):1377-86
pubmed: 18491532
Gastrointest Endosc. 2012 Jul;76(1):126-35
pubmed: 22726471
CA Cancer J Clin. 2018 Jul;68(4):250-281
pubmed: 29846947
Dig Dis Sci. 2017 Oct;62(10):2719-2727
pubmed: 28836087
J Clin Oncol. 2016 Jun 20;34(18):2157-64
pubmed: 27138577
Metabolomics. 2019 Dec 5;16(1):4
pubmed: 31807893
Gastroenterology. 1997 Jan;112(1):24-8
pubmed: 8978338
BMC Gastroenterol. 2015 Apr 01;15:40
pubmed: 25888375
Gut. 2011 Sep;60(9):1236-41
pubmed: 21402615
Gastrointest Endosc. 2007 Jun;65(7):1015-22
pubmed: 17531636
Clin Gastroenterol Hepatol. 2012 Jun;10(6):633-8
pubmed: 22426085
Dig Liver Dis. 2017 Jan;49(1):34-37
pubmed: 27443490
Endoscopy. 2017 Mar;49(3):270-297
pubmed: 28212588
Eur Heart J. 2019 Apr 21;40(16):1268-1276
pubmed: 30615112
Nat Biomed Eng. 2018 Oct;2(10):749-760
pubmed: 31001455
Front Big Data. 2020 Mar 10;3:6
pubmed: 33693381
Clin Cancer Res. 2005 May 15;11(10):3642-6
pubmed: 15897559
Mutat Res Rev Mutat Res. 2019 Jan - Mar;779:45-57
pubmed: 31097151
CA Cancer J Clin. 2018 Nov;68(6):394-424
pubmed: 30207593