Automated surgical action recognition and competency assessment in laparoscopic cholecystectomy: a proof-of-concept study

被引:0
|
作者
Yen, Hung-Hsuan [1 ,2 ]
Hsiao, Yi-Hsiang [1 ,2 ]
Yang, Meng-Han [3 ]
Huang, Jia-Yuan [4 ]
Lin, Hsu-Ting [5 ]
Huang, Chun-Chieh [1 ,2 ]
Blue, Jakey [4 ]
Ho, Ming-Chih [1 ,2 ]
机构
[1] Natl Taiwan Univ Hosp, Dept Surg, Hsin Chu Branch, 2,Sec 1,Shengyi Rd, Zhubei City 302058, Hsinchu County, Taiwan
[2] Natl Taiwan Univ Hosp, Dept Surg, Taipei, Taiwan
[3] Natl Taiwan Univ, Master Program Stat, Taipei, Taiwan
[4] Natl Taiwan Univ, Inst Ind Engn, Taipei, Taiwan
[5] ACE Biotek Co Ltd, Zhubei, Hsinchu County, Taiwan
来源
SURGICAL ENDOSCOPY AND OTHER INTERVENTIONAL TECHNIQUES | 2025年
关键词
Cholecystectomy; Surgical action recognition; Machine learning; Artificial intelligence; Competency assessment; ACUTE CHOLECYSTITIS;
D O I
10.1007/s00464-025-11663-y
中图分类号
R61 [外科手术学];
学科分类号
摘要
BackgroundLaparoscopic cholecystectomy (LC) is a common procedure with standardized steps and validated assessment tools. However, the role of surgical actions in competency assessment remains underexplored, and automated models for surgical action recognition are lacking.MethodsThe Cholec80 dataset of 80 LC videos was analyzed for the Calot's Triangle Dissection (CTD) phase. Strasberg's critical view of safety (CVS) score and second-by-second annotations of surgical actions were evaluated. Videos were categorized into high_simple, low_simple, and high_complex groups based on competency levels and cholecystitis grade. The dataset was randomly divided into training (66 videos) and testing (14 videos) sets based on subgrouping. Surgical metrics were compared between subgroups, and a Random Forest model was constructed to predict competency levels using these metrics. In addition, a Video-Masked Autoencoders (VideoMAE) model was developed for surgical action recognition.ResultsThe high_simple group had significantly shorter CTD duration, fewer action transitions, and lower percentages of suctioning/irrigating, coagulating, and idle actions, but higher CVS scores and dissecting percentages. The Random Forest model achieved 93% accuracy (AUC: 0.96) in competency prediction, with CVS score, CTD duration, and percentages of dissecting, coagulating, and exposing as the top five important features. The VideoMAE model attained 89.11% overall accuracy in recognizing surgical actions, with the highest recall (0.97) for dissecting and the lowest (0.51) for suctioning/irrigating.ConclusionsThis study highlights the importance of surgical actions in competency assessment and presents automated models for evaluation and action recognition. These tools have potential to transform surgical education by providing objective and data-driven feedback for skill improvement.
引用
收藏
页码:3006 / 3016
页数:11
相关论文
共 50 条
  • [41] Spesolimab for hidradenitis suppurativa: A proof-of-concept study
    Alavi, Afsaneh
    Prens, Errol
    Kimball, Alexa B.
    Krueger, James G.
    Mukhopadhyay, Sutirtha
    Wang, Hui
    Ivanoff, Nathalie B.
    Daly, Ana C. Hernandez
    Zouboulis, Christos C.
    JOURNAL OF THE AMERICAN ACADEMY OF DERMATOLOGY, 2023, 89 (03) : AB89 - AB89
  • [42] Proton Grid Therapy: A Proof-of-Concept Study
    Henry, Thomas
    Ureba, Ana
    Valdman, Alexander
    Siegbahn, Albert
    TECHNOLOGY IN CANCER RESEARCH & TREATMENT, 2017, 16 (06) : 749 - 757
  • [43] Network coupling and surgical treatment response in temporal lobe epilepsy: A proof-of-concept study
    Chang, Allen J.
    Roth, Rebecca W.
    Gong, Ruxue
    Gross, Robert E.
    Harmsen, Irene
    Parashos, Alexandra
    Revell, Andrew
    Davis, Kathryn A.
    Bonilha, Leonardo
    Gleichgerrcht, Ezequiel
    EPILEPSY & BEHAVIOR, 2023, 149
  • [44] Use of Google Glass to Enhance Surgical Education of Neurosurgery Residents: "Proof-of-Concept" Study
    Nakhla, Jonathan
    Kobets, Andrew
    Ramos, Rafeal De la Garza
    Haranhalli, Neil
    Gelfand, Yaroslav
    Ammar, Adam
    Echt, Murray
    Scoco, Aleka
    Kinon, Merritt
    Yassari, Reza
    WORLD NEUROSURGERY, 2017, 98 : 711 - 714
  • [45] Simultaneous laparoscopic colectomy and liver metastasectomy with natural orifice specimen extraction: A proof-of-concept study
    Seow-En, Isaac
    Koh, Ye Xin
    Tan, Emile Kwong-Wei
    Tan, Ek Khoon
    HELIYON, 2024, 10 (12)
  • [46] Smart Mobility Digital Twin for Automated Driving: Design and Proof-of-Concept
    Wang, Kui
    Li, Zongdian
    Yu, Tao
    Sakaguchi, Kei
    2023 IEEE 97TH VEHICULAR TECHNOLOGY CONFERENCE, VTC2023-SPRING, 2023,
  • [47] Automated immunohistochemistry of intra-epidermal nerve fibres in skin biopsies: A proof-of-concept study
    Burgess, Jamie
    Marshall, Anne
    Rapteas, Leandros
    Kevin, J. Hamill
    Marshall, Andrew
    Malik, Rayaz A.
    Frank, Bernhard
    Alam, Uazman
    JOURNAL OF THE PERIPHERAL NERVOUS SYSTEM, 2024, 29 (03) : 329 - 338
  • [48] Automated Sequential Injection-Capillary Electrophoresis for DriedBlood Spot Analysis: A Proof-of-Concept Study
    Dvorak, Milos
    Miro, Manuel
    Kuban, Pavel
    ANALYTICAL CHEMISTRY, 2022, 94 (13) : 5301 - 5309
  • [49] Automated detection of focal seizures using subcutaneously implanted electrocardiographic device: A proof-of-concept study
    Jeppesen, Jesper
    Christensen, Jakob
    Molgaard, Henning
    Beniczky, Sandor
    EPILEPSIA, 2023, : S59 - S64
  • [50] An automated diagnostic device for children under 5 years of age: A proof-of-concept study in Malawi
    Shah, M.
    Jenda, G.
    Nsona, H.
    Gigi, E.
    Dadlani, P.
    Swedberg, E.
    EUROPEAN JOURNAL OF PUBLIC HEALTH, 2019, 29