Deep learning for sensitive detection of Helicobacter Pylori in gastric biopsies.
Artificial intelligence
Convolutional neural networks
Deep learning
Gastric cancer prevention
Helicobacter pylori
Screening
Journal
BMC gastroenterology
ISSN: 1471-230X
Titre abrégé: BMC Gastroenterol
Pays: England
ID NLM: 100968547
Informations de publication
Date de publication:
11 Dec 2020
11 Dec 2020
Historique:
received:
02
12
2019
accepted:
08
10
2020
entrez:
14
12
2020
pubmed:
15
12
2020
medline:
15
5
2021
Statut:
epublish
Résumé
Helicobacter pylori, a 2 × 1 μm spiral-shaped bacterium, is the most common risk factor for gastric cancer worldwide. Clinically, patients presenting with symptoms of gastritis, routinely undergo gastric biopsies. The following histo-morphological evaluation dictates therapeutic decisions, where antibiotics are used for H. pylori eradication. There is a strong rational to accelerate the detection process of H. pylori on histological specimens, using novel technologies, such as deep learning. We designed a deep-learning-based decision support algorithm that can be applied on regular whole slide images of gastric biopsies. In detail, we can detect H. pylori both on Giemsa- and regular H&E stained whole slide images. With the help of our decision support algorithm, we show an increased sensitivity in a subset of 87 cases that underwent additional PCR- and immunohistochemical testing to define a sensitive ground truth of HP presence. For Giemsa stained sections, the decision support algorithm achieved a sensitivity of 100% compared to 68.4% (microscopic diagnosis), with a tolerable specificity of 66.2% for the decision support algorithm compared to 92.6 (microscopic diagnosis). Together, we provide the first evidence of a decision support algorithm proving as a sensitive screening option for H. pylori that can potentially aid pathologists to accurately diagnose H. pylori presence on gastric biopsies.
Sections du résumé
BACKGROUND
BACKGROUND
Helicobacter pylori, a 2 × 1 μm spiral-shaped bacterium, is the most common risk factor for gastric cancer worldwide. Clinically, patients presenting with symptoms of gastritis, routinely undergo gastric biopsies. The following histo-morphological evaluation dictates therapeutic decisions, where antibiotics are used for H. pylori eradication. There is a strong rational to accelerate the detection process of H. pylori on histological specimens, using novel technologies, such as deep learning.
METHODS
METHODS
We designed a deep-learning-based decision support algorithm that can be applied on regular whole slide images of gastric biopsies. In detail, we can detect H. pylori both on Giemsa- and regular H&E stained whole slide images.
RESULTS
RESULTS
With the help of our decision support algorithm, we show an increased sensitivity in a subset of 87 cases that underwent additional PCR- and immunohistochemical testing to define a sensitive ground truth of HP presence. For Giemsa stained sections, the decision support algorithm achieved a sensitivity of 100% compared to 68.4% (microscopic diagnosis), with a tolerable specificity of 66.2% for the decision support algorithm compared to 92.6 (microscopic diagnosis).
CONCLUSION
CONCLUSIONS
Together, we provide the first evidence of a decision support algorithm proving as a sensitive screening option for H. pylori that can potentially aid pathologists to accurately diagnose H. pylori presence on gastric biopsies.
Identifiants
pubmed: 33308189
doi: 10.1186/s12876-020-01494-7
pii: 10.1186/s12876-020-01494-7
pmc: PMC7731757
doi:
Types de publication
Journal Article
Langues
eng
Sous-ensembles de citation
IM
Pagination
417Subventions
Organisme : Else Kröner-Fresenius-Stiftung
ID : 2015_Kolleg.19
Références
Cochrane Database Syst Rev. 2015 Jul 22;(7):CD005583
pubmed: 26198377
Endosc Int Open. 2018 Feb;6(2):E139-E144
pubmed: 29399610
CA Cancer J Clin. 2019 Jan;69(1):7-34
pubmed: 30620402
BMJ. 1991 Jun 1;302(6788):1302-5
pubmed: 2059685
J Infect Dis. 1993 Aug;168(2):379-85
pubmed: 8335974
JAMA. 2013 Apr 3;309(13):1351-2
pubmed: 23549579
N Engl J Med. 1995 Jul 6;333(1):32-41
pubmed: 7776992
Nat Rev Gastroenterol Hepatol. 2017 Apr;14(4):230-240
pubmed: 28053340
Am J Gastroenterol. 2007 Aug;102(8):1808-25
pubmed: 17608775
EBioMedicine. 2017 Nov;25:106-111
pubmed: 29056541
N Engl J Med. 1991 Oct 17;325(16):1132-6
pubmed: 1891021
Gut. 2012 May;61(5):646-64
pubmed: 22491499
Lancet. 2019 May 11;393(10184):1948-1957
pubmed: 30982686
Arch Pathol Lab Med. 2020 Mar;144(3):370-378
pubmed: 31246112
Endoscopy. 2004 Jul;36(7):601-8
pubmed: 15243882
Am J Gastroenterol. 2005 May;100(5):1037-42
pubmed: 15842576
Cancer Cell. 2018 Jan 8;33(1):137-150.e5
pubmed: 29290541
Nature. 2015 May 28;521(7553):436-44
pubmed: 26017442
Gastroenterology. 2009 Nov;137(5):1641-8.e1-2
pubmed: 19664631
IARC Monogr Eval Carcinog Risks Hum. 1994;61:1-241
pubmed: 7715068
N Engl J Med. 1991 Oct 17;325(16):1127-31
pubmed: 1891020