Computer Vision Analysis of Intraoperative Video: Automated Recognition of Operative Steps in Laparoscopic Sleeve Gastrectomy.
Academic Medical Centers
Adult
Artificial Intelligence
Automation
Databases, Factual
Female
Gastrectomy
/ methods
Humans
Laparoscopy
/ methods
Male
Middle Aged
Monitoring, Intraoperative
/ methods
Observer Variation
Operative Time
Retrospective Studies
Sensitivity and Specificity
Video Recording
/ statistics & numerical data
Video-Assisted Surgery
/ methods
Journal
Annals of surgery
ISSN: 1528-1140
Titre abrégé: Ann Surg
Pays: United States
ID NLM: 0372354
Informations de publication
Date de publication:
09 2019
09 2019
Historique:
pubmed:
6
7
2019
medline:
13
3
2020
entrez:
6
7
2019
Statut:
ppublish
Résumé
To develop and assess AI algorithms to identify operative steps in laparoscopic sleeve gastrectomy (LSG). Computer vision, a form of artificial intelligence (AI), allows for quantitative analysis of video by computers for identification of objects and patterns, such as in autonomous driving. Intraoperative video from LSG from an academic institution was annotated by 2 fellowship-trained, board-certified bariatric surgeons. Videos were segmented into the following steps: 1) port placement, 2) liver retraction, 3) liver biopsy, 4) gastrocolic ligament dissection, 5) stapling of the stomach, 6) bagging specimen, and 7) final inspection of staple line. Deep neural networks were used to analyze videos. Accuracy of operative step identification by the AI was determined by comparing to surgeon annotations. Eighty-eight cases of LSG were analyzed. A random 70% sample of these clips was used to train the AI and 30% to test the AI's performance. Mean concordance correlation coefficient for human annotators was 0.862, suggesting excellent agreement. Mean (±SD) accuracy of the AI in identifying operative steps in the test set was 82% ± 4% with a maximum of 85.6%. AI can extract quantitative surgical data from video with 85.6% accuracy. This suggests operative video could be used as a quantitative data source for research in intraoperative clinical decision support, risk prediction, or outcomes studies.
Sections du résumé
OBJECTIVE(S)
To develop and assess AI algorithms to identify operative steps in laparoscopic sleeve gastrectomy (LSG).
BACKGROUND
Computer vision, a form of artificial intelligence (AI), allows for quantitative analysis of video by computers for identification of objects and patterns, such as in autonomous driving.
METHODS
Intraoperative video from LSG from an academic institution was annotated by 2 fellowship-trained, board-certified bariatric surgeons. Videos were segmented into the following steps: 1) port placement, 2) liver retraction, 3) liver biopsy, 4) gastrocolic ligament dissection, 5) stapling of the stomach, 6) bagging specimen, and 7) final inspection of staple line. Deep neural networks were used to analyze videos. Accuracy of operative step identification by the AI was determined by comparing to surgeon annotations.
RESULTS
Eighty-eight cases of LSG were analyzed. A random 70% sample of these clips was used to train the AI and 30% to test the AI's performance. Mean concordance correlation coefficient for human annotators was 0.862, suggesting excellent agreement. Mean (±SD) accuracy of the AI in identifying operative steps in the test set was 82% ± 4% with a maximum of 85.6%.
CONCLUSIONS
AI can extract quantitative surgical data from video with 85.6% accuracy. This suggests operative video could be used as a quantitative data source for research in intraoperative clinical decision support, risk prediction, or outcomes studies.
Identifiants
pubmed: 31274652
doi: 10.1097/SLA.0000000000003460
pmc: PMC7216040
mid: NIHMS1585507
doi:
Types de publication
Journal Article
Research Support, N.I.H., Extramural
Research Support, Non-U.S. Gov't
Langues
eng
Sous-ensembles de citation
IM
Pagination
414-421Subventions
Organisme : NIDDK NIH HHS
ID : T32 DK007754
Pays : United States
Références
IEEE Trans Med Imaging. 2017 Jan;36(1):86-97
pubmed: 27455522
IEEE Trans Med Imaging. 2015 Apr;34(4):877-87
pubmed: 25373078
JAMA Surg. 2019 May 1;154(5):389-390
pubmed: 30673061
J Biomed Inform. 2011 Jun;44(3):486-96
pubmed: 20096376
Neural Comput. 1997 Nov 15;9(8):1735-80
pubmed: 9377276
J Minim Invasive Gynecol. 2017 Mar - Apr;24(3):426-431
pubmed: 28063907
Mil Med. 2013 Oct;178(10 Suppl):22-7
pubmed: 24084302
JAMA Surg. 2017 Apr 1;152(4):318-325
pubmed: 27973648
J Surg Educ. 2018 Nov;75(6):1475-1479
pubmed: 29699931
Int J Comput Assist Radiol Surg. 2017 Feb;12(2):325-338
pubmed: 27573276
J Biomed Inform. 2015 Apr;54:158-66
pubmed: 25752728
Am J Surg. 2019 Feb;217(2):356-361
pubmed: 30470551
Ann Surg. 2018 Jul;268(1):70-76
pubmed: 29389679
Ann Surg. 2017 Dec;266(6):934-936
pubmed: 28338514
Surg Endosc. 2017 Dec;31(12):5094-5100
pubmed: 28444497
J Surg Educ. 2017 Mar - Apr;74(2):365-371
pubmed: 27720404
Ann Surg. 2018 May;267(5):868-873
pubmed: 28650360
JAMA Surg. 2019 May 1;154(5):381-389
pubmed: 30673072
Nat Med. 2019 Jan;25(1):44-56
pubmed: 30617339