Revista do Colégio Brasileiro de Cirurgiões, Journal Year: 2023, Volume and Issue: 50
Published: Jan. 1, 2023
Frontiers in Surgery, Journal Year: 2025, Volume and Issue: 12
Published: Feb. 26, 2025
This systematic literature review of the integration artificial intelligence (AI) applications in surgical practice through hand and instrument tracking provides an overview recent advancements analyzes current on intersection surgery with AI. Distinct AI algorithms specific are also examined. An advanced search using medical subject heading terms was conducted Medline (via PubMed), SCOPUS, Embase databases for articles published English. A strict selection process performed, adhering to PRISMA guidelines. total 225 were retrieved. After screening, 77 met inclusion criteria included review. Use uncommon during 2013-2017 but has gained significant popularity since 2018. Deep learning (n = 62) increasingly preferred over traditional machine 15). These technologies used fields such as general 19), neurosurgery 10), ophthalmology 9). The most common functional sensors systems prerecorded videos 29), cameras 21), image datasets 7). laparoscopic 13), robotic-assisted basic 12), endoscopic 8) skills training, well simulation training 8). can be tailored address distinct needs education patient care. use improves outcomes by optimizing training. It is essential acknowledge technical social limitations work toward filling those gaps future studies.
Language: Английский
Citations
1
Surgical Endoscopy, Journal Year: 2023, Volume and Issue: 37(11), P. 8818 - 8828
Published: Aug. 25, 2023
Abstract Introduction Artificial intelligence and computer vision are revolutionizing the way we perceive video analysis in minimally invasive surgery. This emerging technology has increasingly been leveraged successfully for segmentation, documentation, education, formative assessment. New, sophisticated platforms allow pre-determined segments chosen by surgeons to be automatically presented without need review entire videos. study aimed validate demonstrate accuracy of first reported AI-based algorithm that recognizes surgical steps videos totally extraperitoneal (TEP) inguinal hernia repair. Methods Videos TEP procedures were manually labeled a team annotators trained identify label workflow according six major steps. For bilateral hernias, an additional change focus step was also included. The then used train AI algorithm. Performance assessed comparison manual annotations. Results A total 619 full-length analyzed: 371 model, 93 internal validation, remaining 155 as test set evaluate accuracy. overall complete procedure 88.8%. Per-step reached highest value sac reduction (94.3%) lowest preperitoneal dissection (72.2%). Conclusions These results indicate novel model able provide fully automated with high level. High-accuracy models leveraging enable automation us monitor performance, providing mathematical metrics can stored, evaluated, compared. As such, proposed is capable enabling data-driven insights improve quality best practices procedures. Graphical abstract
Language: Английский
Citations
17
Surgical Endoscopy, Journal Year: 2023, Volume and Issue: 37(11), P. 8690 - 8707
Published: July 29, 2023
Abstract Background Surgery generates a vast amount of data from each procedure. Particularly video provides significant value for surgical research, clinical outcome assessment, quality control, and education. The lifecycle is influenced by various factors, including structure, acquisition, storage, sharing; use exploration, finally governance, which encompasses all ethical legal regulations associated with the data. There universal need among stakeholders in science to establish standardized frameworks that address aspects this ensure purpose. Methods Working groups were formed, 48 representatives academia industry, clinicians, computer scientists industry representatives. These working focused on: Data Use, Structure, Exploration, Governance. After group panel discussions, modified Delphi process was conducted. Results resulting consensus conceptualized structured recommendations domain related We identified key within formulated comprehensive, easily understandable, widely applicable guidelines utilization. Standardization structure should encompass format quality, sources, documentation, metadata, account biases To foster scientific datasets reflect diversity remain adaptable future applications. governance must be transparent stakeholders, addressing considerations surrounding Conclusion This presents essential around generation diverse databanks, accounting multiple involved throughout its lifecycle. Following SAGES annotation framework, we lay foundation standardization use, exploration. A detailed exploration requirements adequate will follow.
Language: Английский
Citations
15
EClinicalMedicine, Journal Year: 2024, Volume and Issue: 70, P. 102545 - 102545
Published: March 22, 2024
BackgroundSurgical video contains data with significant potential to improve surgical outcome assessment, quality assurance, education, and research. Current utilisation of recording is unknown related policies/governance structures are unclear.MethodsA nationwide Freedom Information (FOI) request concerning recording, technology, consent, access, governance was sent all acute National Health Service (NHS) trusts/boards in England/Wales between 20th February March 2023.Findings140/144 (97.2%) responded the FOI request. Surgical procedures were routinely recorded 22 trusts/boards. The median estimate consultant surgeons their 20%. stored on internal systems (n = 27), third-party products 29), both 9). 32/140 (22.9%) ask for consent record as part routine care. Consent included non-clinical purposes 55/140 (39.3%) Policies surgeon/patient access available 48/140 (34.3%) trusts/boards, respectively. used 64/140 (45.7%) Governance policies covering use, and/or storage from 59/140 (42.1%) trusts/boards.InterpretationThere heterogeneity practices England Wales. A minority procedures, large variation recording/storage indicating scope NHS-wide coordination. Revision accessibility, should be prioritised by protect key stakeholders. Increased availability essential patients maximally benefit ongoing digital transformation surgery.FundingKL supported an NIHR Academic Clinical Fellowship acknowledges infrastructure support this research Institute Research (NIHR) Imperial Biomedical Centre (BRC).
Language: Английский
Citations
5
Global Surgical Education - Journal of the Association for Surgical Education, Journal Year: 2024, Volume and Issue: 3(1)
Published: May 10, 2024
Abstract There are numerous barriers in robotic surgical training, including reliance on observational learning, low-quality feedback, and inconsistent assessment. Artificial intelligence (AI) offers potential solutions to these central problems education may allow for more efficient efficacious training. Three key areas which AI has particular relevance video labeling, Video labeling refers the automated designation of prespecified categories operative videos. Numerous prior studies have applied particularly retrospective educational review after an operation. allows learners their instructors rapidly identify critical parts video. We recommend incorporating AI-based into where available. also a mechanism by reliable feedback can be provided surgery. Feedback through harnesses performance metrics (APMs) natural language processing (NLP) provide actionable descriptive plans while reducing faculty assessment burden. combining supervised AI-generated, APM-based with expert-based surgeons trainees reflect like bimanual dexterity efficiency. Finally, summative could appraisal or trainees. However, remains limited concerns around bias opaque processes. Several computer vision compare expert-completed rating scales, though such work investigational. At this time, we against use pending additional validity evidence. Overall, promising future directions address multiple challenges Through advances assessment, demonstrated ways increase efficiency efficacy education.
Language: Английский
Citations
4
Global Surgical Education - Journal of the Association for Surgical Education, Journal Year: 2025, Volume and Issue: 4(1)
Published: Feb. 1, 2025
Language: Английский
Citations
0
Expert Review of Gastroenterology & Hepatology, Journal Year: 2024, Volume and Issue: 18(1-3), P. 89 - 102
Published: March 3, 2024
Introduction Strategies to promote high-quality endoscopy in children require consensus around pediatric-specific quality standards and indicators. Using a rigorous guideline development process, the international Pediatric Endoscopy Quality Improvement Network (PEnQuIN) was developed support continuous improvement efforts within across pediatric services.
Language: Английский
Citations
1
Surgical Practice, Journal Year: 2023, Volume and Issue: 27(3), P. 128 - 130
Published: Aug. 1, 2023
Citations
2
Revista do Colégio Brasileiro de Cirurgiões, Journal Year: 2023, Volume and Issue: 50
Published: Jan. 1, 2023
Citations
2
Global Surgical Education - Journal of the Association for Surgical Education, Journal Year: 2024, Volume and Issue: 3(1)
Published: Aug. 1, 2024
Abstract Surgical Data Science is an emerging scientific discipline that applies advances in data science, machine learning and AI to harness the increasingly large amounts of surgical enable surgery [1–4]. collection for solutions involves both ingestion contingent (in case surgery—medical records, data, instrument medical images, from OR sensors video), as well intentionally collected annotations expert opinion describing data. This organized knowledge then used train models ultimately generate predictions based on available training Historically, science workflow starts with organizing a clean consistent dataset, mantra GIGO— garbage in, out —emphasizing quality model output directly related In surgery, healthcare general, this not easy goal achieve due complex logistics collection, missing incomplete human error, lack measurement standards, subjective differences interpretation. article, we look at particular perspective uncertainty. We highlight few topics which hospitals, surgeons research teams need be aware when collecting will provide actionable outputs clinical educational settings.
Language: Английский
Citations
0