POSSIBILITIES OF ARTIFICIAL INTELLIGENCE IN THE CHOICE OF A THERAPEUTIC AND DIAGNOSTIC ALGORITHM FOR BLEEDING FROM THE GASTROINTESTINAL TRACT

DOI: 10.22138/2500-0918-2023-20-3-109-120

A.A. Zhilyakov, S.Yu. Sokolov, S.A. Chernyadyev,
O.V. Kirshina, N.Yu. Korobova

Federal State Budgetary Educational Institution of Higher Education Ural State Medical University of the Ministry of Health of the Russian Federation, Yekaterinburg, Russian Federation

Abstract. The purpose of this review is to analyze the current literature on the use of artificial intelligence in the diagnosis of ulcerative bleeding, in order to assess the current status and formulate hypotheses regarding the further development of this technology and the requirements for its integration into practice. Materials and methods: The study used articles from the PubMed database, selected for the keywords «artificial intelligence», «ulcerative bleeding», «neurotic bleeding», «deep learning» or «machine learning». Results: The study of the literature made it possible to establish that the use of modern methods of treatment reduced the frequency of ulcerative bleeding in patients and the percentage of errors in endoscopists. However, the treatment of such patients requires the joint work of doctors of several specialties. The use of artificial intelligence (AI) significantly improves the quality of doctors’ work and, in particular, AI systems have shown great potential in many areas of gastroenterology. Recently, an AI-powered system for detecting and diagnosing polyps by colonoscopy was introduced, which improves the quality of endoscopists’ work. Conclusion: In this regard, the use of artificial intelligence technologies is a promising direction in the treatment of ulcerative bleeding and can significantly improve the practice of endoscopists

Keywords: peptic ulcer; bleeding; deep training; artificial intelligence

There is no conflict of interest.
Contact details of the corresponding author:
Alexander A. Zhilyakov
usma@usma.ru
Received 12.01.2023
For citation:
Zhilyakov A.A., Sokolov S.Yu., Chernyadyev S.A., Kirshina O.V., Korobova N.Yu. Possibilities of artificial intelligence in the choice of a therapeutic and diagnostic algorithm for bleeding from the gastrointestinal tract. [Online] Vestn. Ural. Med. Akad. Nauki. = Journal of Ural Medical Academic Science. 2023, Vol. 20, no. 3, pp. 109–120. DOI: 10.22138/2500-0918-2023-20-3-109-120 (In Russ)

 
REFERENCES
1. Bello F.P.S. et al. Acute upper gastrointestinal bleeding due to portal hypertension in children: What is the best timing of endoscopy? Dig Liver Dis, 2022 Vol. 54 № 1. P. 63-68. doi: 10.1016/j.dld.2021.09.010.
2. Segal J. et al. Training in endotherapy for acute upper gastrointestinal bleeding: a UK-wide gastroenterology trainee survey. Frontline Gastroenterol, 2020. Vol. 11, № 6. P. 430-435. doi: 10.1136/flgastro-2019-101345.
3. Gralnek I.M. et al. Endoscopic diagnosis and management of nonvariceal upper gastrointestinal hemorrhage (NVUGIH): European Society of Gastrointestinal Endoscopy (ESGE) Guideline — Update 2021. Endoscopy, 2021. Vol. 53, № 3. P. 300-332. doi: 10.1055/a-1369-5274.
4. Lau L.H.S., Sung J.J.Y. Treatment of upper gastrointestinal bleeding in 2020: New techniques and outcomes. Dig Endosc. Blackwell Publishing, 2021. Vol. 33, № 1. P. 83–94. doi:10.1111/den.13674
5. Shi W. et al. China Acute Myocardial Infarction (CAMI) Registry Study Group. In-hospital gastrointestinal bleeding in patients with acute myocardial infarction: incidence, outcomes and risk factors analysis from China Acute Myocardial Infarction Registry. BMJ Open, 2021. Vol. 11, № 9. P. e044117. doi: 10.1136/bmjopen-2020-044117.
6. Weissman S. et al. Relationships of hospitalization outcomes and timing to endoscopy in non-variceal upper gastrointestinal bleeding: A nationwide analysis. World J Gastrointest Endosc, 2023. Vol. 15, № 4. P. 285-296. doi: 10.4253/wjge.v15.i4.285.
7. Mullady D.K. et al. AGA Clinical Practice Update on Endoscopic Therapies for Non-Variceal Upper Gastrointestinal Bleeding: Expert Review. Gastroenterology, 2020. Vol. 159, № 3. P. 1120-1128. doi: 10.1053/j.gastro.2020.05.095.
8. Bordin D.C. et al. Helicobacter pylori: clinical significance and principles of diagnosis. Infectious diseases: news, opinions, training. 2022. Vol. 11, № 1. P. 119–129. doi:10.33029/2305-3496-2022-11-1-119-129
9. Pulatov S.A., Motyko A.A. Development of an algorithm for automatic detection of bleeding on endoscopic images. Science of the present and the future. 2019. Vol. 1, P. 185–187. https://www.elibrary.ru/item.asp?id=41177370
10. Alali A.A. et al. Topical hemostatic agents in the management of upper gastrointestinal bleeding: a meta-analysis. Endosc Int Open, 2023. Vol. 11, № 4. P. E368-E385. doi: 10.1055/a-1984-6895.
11. Shung D. et al. Early identification of patients with acute gastrointestinal bleeding using natural language processing and decision rules. Journal of Gastroenterology and Hepatology (Australia). John Wiley and Sons Inc, 2021. Vol. 36, № 6. P. 1590–1597. doi:10.1111/JGH.15313
12. Alali A.A., Barkun A.N. An update on the management of non-variceal upper gastrointestinal bleeding. Gastroenterol Rep (Oxf), 2023. Vol. 11:goad011. doi: 10.1093/gastro/goad011.
13. Vorobyov P.A. Theory in the construction of automated diagnostic questionnaires // Problems of standardization in health care. 2016. Vol. 7, № 8. P. 3–13. https://www.elibrary.ru/item.asp?id=26562019
14. Rockall T.A. et al. Risk assessment after acute upper gastrointestinal haemorrhage. Gut. BMJ Publishing Group, 1996. Vol. 38, № 3. P. 316–321. doi:10.1136/GUT.38.3.316
15. Kim M.S., Choi J., Shin W.C. AIMS65 scoring system is comparable to Glasgow-Blatchford score or Rockall score for prediction of clinical outcomes for non-variceal upper gastrointestinal bleeding. BMC Gastroenterol. BioMed Central Ltd., 2019. Vol. 19, № 1. doi:10.1186/S12876-019-1051-8
16. Ebrahimi Bakhtavar H. et al. Clinical Scoring Systems in Predicting the Outcome of Acute Upper Gastrointestinal Bleeding; a Narrative Review. Emerg (Tehran), 2017. Vol. 5, № 1. P. e36.
17. Park S.M. et al. Comparison of AIMS65 Score and Other Scoring Systems for Predicting Clinical Outcomes in Koreans with Nonvariceal Upper Gastrointestinal Bleeding. Gut Liver, 2016. Vol. 10, № 4. P. 526-31. doi: 10.5009/gnl15153.
18. Shung D.L. et al. Validation of a Machine Learning Model That Outperforms Clinical Risk Scoring Systems for Upper Gastrointestinal Bleeding. Gastroenterology, 2020. Vol. 158, № 1. P. 160-167. doi: 10.1053/j.gastro.2019.09.009.
19. Shung D.L. et al. Validation of a Machine Learning Model That Outperforms Clinical Risk Scoring Systems for Upper Gastrointestinal Bleeding. Gastroenterology. W.B. Saunders, 2020. Vol. 158, № 1. P. 160–167. doi:10.1053/J.GASTRO.2019.09.009
20. Berbís M.A. et al. A. Role of artificial intelligence in multidisciplinary imaging diagnosis of gastrointestinal diseases. World J Gastroenterol, 2021. Vol. 27, № 27. P. 4395-4412. doi: 10.3748/wjg.v27.i27.4395.
21. Kang S. et al. Machine Learning Model for the Prediction of Hemorrhage in Intensive Care Units. Healthc Inform Res, 2022. Vol. 28, № 4. P. 364-375. doi: 10.4258/hir.2022.28.4.364.
22. Forrest J.A.H., Finlayson N.D.C., Shearman D.J.C. Endoscopy in gastrointestinal bleeding. The Lancet. 1974. Vol. 304, № 7877. P. 394–397. doi:10.1016/S0140-6736, № 74)91770-X
23. Yen H.H. et al. Performance Comparison of the Deep Learning and the Human Endoscopist for Bleeding Peptic Ulcer Disease. J Med Biol Eng. Springer Science and Business Media Deutschland GmbH, 2021. Vol. 41, № 4. P. 504–513. doi:10.1007/S40846-021-00608-0
24. Lai Y. et al. Development and validation of a model to predict rebleeding within three days after endoscopic hemostasis for high-risk peptic ulcer bleeding. BMC Gastroenterol, 2022. Vol. 22, № 1. P. 64. doi: 10.1186/s12876-022-02145-9.
25. Laine L., Freeman M., Cohen H. Lack of uniformity in evaluation of endoscopic prognostic features of bleeding ulcers. Gastrointest Endosc. 1994. Vol. 40, № 4. P. 411–417. doi:10.1016/S0016-5107(94)70202-0
26. Yen H.H. et al. Forrest Classification for Bleeding Peptic Ulcer: A New Look at the Old Endoscopic Classification. Diagnostics (Basel), 2022. Vol. 12 № 5. P. 1066. doi: 10.3390/diagnostics12051066.
27. Deshmukh F., Merchant S.S. Explainable Machine Learning Model for Predicting GI Bleed Mortality in the Intensive Care Unit. American Journal of Gastroenterology. Wolters Kluwer Health, 2020. Vol. 115, № 10. P. 1657–1668. doi:10.14309/AJG.0000000000000632
28. Khasanov A.G., Nurtdinov M.A., Gololobov G.Yu. About predicting complications of peptic ulcer disease based on artificial neural networks. Annals of Surgery. 2016. Vol. 21, №. 4. P. 231-234. doi:10.18821/1560-9502-2016-21-4-231-234
29. Sun C. et al. Interpretable time-aware and co-occurrence-aware network for medical prediction. BMC Med Inform Decis Mak, 2021. Vol. 21, № 1. P. 305. doi: 10.1186/s12911-021-01662-z.
30. Dilaghi E. et al. Systematic review and meta-analysis: Artificial intelligence for the diagnosis of gastric precancerous lesions and Helicobacter pylori infection. Dig Liver Dis. Vol. 54, № 12. P. 1630-1638. doi: 10.1016/j.dld.2022.03.007.
31. Wong G.L.H. et al. Machine learning model to predict recurrent ulcer bleeding in patients with history of idiopathic gastroduodenal ulcer bleeding. Aliment Pharmacol Ther. Blackwell Publishing Ltd, 2019. Vol. 49, № 7. P. 912–918. doi:10.1111/APT.15145
32. Wong G.L.H. et al. Prevention of recurrent idiopathic gastroduodenal ulcer bleeding: A double-blind, randomised trial. Gut. BMJ Publishing Group, 2020. Vol. 69, № 4. P. 652–657. doi:10.1136/GUTJNL-2019-318715
33. Kobrinsky B.A., Khavkin A.I., Volynets G.V. Prospects for the use of artificial intelligence systems in gastroenterology. 2020. Vol. 179, № 7. P. 109–117. doi:10.31146/1682-8658-ecg-179-7-109-117
34. Zhao Q., Chi T. Biopsy in emergency gastroscopy does not increase the risk of rebleeding in patients with Forrest I acute nonvariceal upper gastrointestinal bleeding combined with suspected malignant gastric ulcer: a multicenter retrospective cohort study. BMC Gastroenterol. BioMed Central Ltd, 2021. Vol. 21, № 1. doi:10.1186/S12876-021-01836-Z
35. Afonso J. et al. Automated detection of ulcers and erosions in capsule endoscopy images using a convolutional neural network. Med Biol Eng Comput. 2022. Vol. 60, № 3. P. 719-725. doi: 10.1007/s11517-021-02486-9.
36. Cho B.J. et al. Automated classification of gastric neoplasms in endoscopic images using a convolutional neural network. Endoscopy. Georg Thieme Verlag, 2019. Vol. 51, № 12. P. 1121–1129. doi:10.1055/A-0981-6133
37. Namikawa K. et al. Artificial intelligence-based diagnostic system classifying gastric cancers and ulcers: Comparison between the original and newly developed systems. Endoscopy. Georg Thieme Verlag, 2020. Vol. 52, № 12. P. 1077–1083. doi:10.1055/A-1194-8771
38. Arkhipova A. A., Anishchenko V. B. Modern possibilities and prospects for early diagnosis of gastric cancer. Acta Biomed Sci. 2021. Vol. 6, № 3. P. 113–125. doi:10.29413 / ABS. 2021-6.3.12
39. Hamada K. et al. Application of convolutional neural networks for evaluating the depth of invasion of early gastric cancer based on endoscopic images. J Gastroenterol Hepatol, 2022. Vol. 37, № 2. P. 352-357. doi: 10.1111/jgh.15725.
40. Yen T. et al. Optimizing Endoscopy Procedure Documentation Improves Guideline-Adherent Care in Upper Gastrointestinal Bleeding. Dig Dis Sci, 2023 № 16. P. 1–12. doi: 10.1007/s10620-023-07823-6.
41. Alatawi A. et al.. Findings of Esophagogastroduodenoscopy in Patients Suspected of Upper Gastrointestinal Bleeding Referred to the Main Endoscopy Unit at King Fahad Specialist Hospital. Cureus, 2020. Vol. 12, № 12. P. e11862. doi: 10.7759/cureus.11862.
42. Thompson W.H. et al. Dataset decay and the problem of sequential analyses on open datasets. Elife. eLife Sciences Publications Ltd, 2020. Vol. 9. P. 1–17. doi:10.7554/ELIFE.53498
43. Smedsrud P.H. et al. Kvasir-Capsule, a video capsule endoscopy dataset. Sci Data, 2021. Vol. 8, № 1. P. 142. doi: 10.1038/s41597-021-00920-z.
44. Yen H.H. et al. Current Status and Future Perspective of Artificial Intelligence in the Management of Peptic Ulcer Bleeding: A Review of Recent Literature. J Clin Med, 2021. Vol. 10, № 16. P. 3527. doi: 10.3390/jcm10163527.
45. Yu Q. et al. Empiric Transcatheter Embolization for Acute Arterial Upper Gastrointestinal Bleeding: A Meta-Analysis. AJR Am J Roentgenol, 2021. Vol. 216, № 4. P. 880-893. doi: 10.2214/AJR.20.23151.
46. Ini’ C. et al. Embolization for acute nonvariceal bleeding of upper and lower gastrointestinal tract: a systematic review. CVIR Endovasc, 2023. Vol. 6, № 1. P. 18. doi: 10.1186/s42155-023-00360-3.
47. Vorčák M. et al. Endovascular Treatment of Gastrointestinal Hemorrhage. Medicina (Kaunas), 2022. Vol. 58, № 3. P. 424. doi: 10.3390/medicina58030424.
48. Yastremsky A.P. et al. Teaching an artificial neural network as a stage in the development of an expert system for differential diagnosis of acute pharyngeal diseases. Ural. Med. Journal, 2015, Vol. 5, №. 128, P. 74–79. https://elibrary.ru/item.asp?id=24313600.

Authors
Alexander A. Zhilyakov
4th year student of the Faculty of Medicine and Prevention

Sergey Yu. Sokolov
Ph.D., Associate Professor, Head of the Department of Medical Physics, Computer Science and Mathematics

Sergey A. Chernyadyev
Doctor of Medical Sciences, Professor, Head of the Department of Surgical Diseases, Cardiovascular Surgery, Reconstructive and Plastic Surgery

Olga V. Kirshina
Doctor of Medical Sciences, Professor, ProfessorDepartment of Surgical Diseases, Cardiovascular Surgery, Reconstructive and Plastic Surgery

Natalya Yu. Korobova
Ph.D., Associate Professor, Assistant Professor, Department of Surgical Diseases, Cardiovascular Surgery, Reconstructive and Plastic Surgery

Federal State Budgetary Educational Institution of Higher Education Ural State Medical University of the Ministry of Health of the Russian Federation
Yekaterinburg, Russian Federation
usma@usma.ru