Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

S4M: 4-points to segment anything.

International journal of computer assisted radiology and surgery·2026
Same author

Comparison of oncological outcomes between the first 28 cases of TaTME and LaTME: A matched case-control study.

Journal of visceral surgery·2026
Same author

Where are they looking in the operating room?

International journal of computer assisted radiology and surgery·2026
Same author

A fully transanal endoscopic approach for large post-anastomotic high rectovaginal fistulas: An IDEAL stage 1 technical note.

Colorectal disease : the official journal of the Association of Coloproctology of Great Britain and Ireland·2026
Same author

Three-dimensional image guidance for diagnosis and treatment of adrenal disease: a systematic review.

Updates in surgery·2025
Same author

Ultrasam: a foundation model for ultrasound using large open-access segmentation datasets.

International journal of computer assisted radiology and surgery·2025

Related Experiment Video

Updated: Aug 6, 2025

Author Spotlight: Revolutionizing Remote Surgery with Augmented Reality and Robotics for Enhanced Precision and Accessibility
07:46

Author Spotlight: Revolutionizing Remote Surgery with Augmented Reality and Robotics for Enhanced Precision and Accessibility

Published on: August 9, 2024

799

TRandAugment: temporal random augmentation strategy for surgical activity recognition from videos.

Sanat Ramesh1,2, Diego Dall'Alba3, Cristians Gonzalez4,5

  • 1Altair Robotics Lab, University of Verona, 37134, Verona, Italy. sanat.ramesh@univr.it.

International Journal of Computer Assisted Radiology and Surgery
|March 22, 2023
PubMed
Summary

A new method, TRandAugment, enhances surgical video analysis by applying consistent random transformations to temporal segments. This automated approach improves surgical phase and step recognition accuracy by 1-6% over existing methods.

Keywords:
Cataract proceduresData augmentationGastric bypass proceduresSurgical activity recognitionTemporal augmentationTemporal convolutional networks

More Related Videos

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
04:48

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

2.8K
Intra-Operative Behavioral Tasks in Awake Humans Undergoing Deep Brain Stimulation Surgery
12:04

Intra-Operative Behavioral Tasks in Awake Humans Undergoing Deep Brain Stimulation Surgery

Published on: January 6, 2011

13.1K

Related Experiment Videos

Last Updated: Aug 6, 2025

Author Spotlight: Revolutionizing Remote Surgery with Augmented Reality and Robotics for Enhanced Precision and Accessibility
07:46

Author Spotlight: Revolutionizing Remote Surgery with Augmented Reality and Robotics for Enhanced Precision and Accessibility

Published on: August 9, 2024

799
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
04:48

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

2.8K
Intra-Operative Behavioral Tasks in Awake Humans Undergoing Deep Brain Stimulation Surgery
12:04

Intra-Operative Behavioral Tasks in Awake Humans Undergoing Deep Brain Stimulation Surgery

Published on: January 6, 2011

13.1K

Area of Science:

  • Medical image analysis
  • Computer-assisted interventions
  • Surgical robotics

Background:

  • Automatic recognition of surgical activities from intraoperative videos is vital for intelligent support systems.
  • Deep learning models for surgical video analysis benefit from data augmentation to improve generalization.
  • Augmenting still images is established, but extending to videos requires addressing the temporal dimension and challenges of long, complex surgical activities.

Purpose of the Study:

  • To introduce TRandAugment, a simplified, automated data augmentation technique specifically for long surgical videos.
  • To address the complexities of temporal data augmentation in surgical video analysis.
  • To enhance the generalization and performance of deep learning models in surgical activity recognition.

Main Methods:

  • TRandAugment treats surgical videos as sequences of temporal segments, applying consistent random transformations within each segment.
  • An end-to-end spatiotemporal model, integrating a Convolutional Neural Network (CNN - ResNet50) with a Temporal Convolutional Network (TCN), was trained using the proposed augmentation method.
  • The method was evaluated on the Bypass40 and CATARACTS surgical video datasets for surgical phase and step recognition tasks.

Main Results:

  • TRandAugment demonstrated effectiveness on two distinct surgical video datasets and two recognition tasks.
  • The proposed method achieved a performance improvement of 1-6% compared to state-of-the-art methods utilizing manual augmentations.
  • This highlights the advantage of automated temporal augmentation over manual strategies for surgical videos.

Conclusions:

  • A simplified and automated augmentation method, TRandAugment, has been developed for long surgical videos.
  • Validation across different datasets and tasks confirms the significance of temporal augmentation strategies for surgical video analysis.
  • The findings underscore the need for specialized temporal augmentation techniques to advance intelligent surgical support systems.