Table of Contents

Automatic Assessment of Surgical Ergonomics

Last updated: 2/25/21 11:59 PM

Summary

For this project our team is attempting to assess surgical ergonomics using MATLAB and video files to ensure that surgeons are performing surgeries with good posture. We use the RULA and ROSA ergonomic assessments to test this.

Background, Specific Aims, and Significance

Surgeries in the OR can result in surgeons adopting awkward postures, which can lead to poor circulation, muscle/join pain, and increased fatigue. Studies have shown that surgeons spend over 50% of the time with their neck in flexion, which is often associated with tension myalgia. Suboptimal postures in general are associated with poor efficiency, and an optimized surgical setting can improve both efficiency and performance in the OR. The aims of this project are to

1. Automate a way for surgeons to see if they are spending times in dangerous posture positions- this saves time and resources so that surgeons can focus on surgery, and can quickly receive feedback about their posture.

2. Give ways for surgeons to improve their posture during surgery and show how much time is spent in dangerous zones- this allows surgeons to understand whether the awkward positions they are in are dangerous for themselves and understand risks to their own bodies.

Deliverables

Technical Approach

We will use C++, an Intel Real Sense D415 RGBD Camera, and a SDK to essentially take in video files as input and convert humans into 3-D stick figures. From there, we can calculate frame by frame the angles between joints on the stick figure, and plug these angles into existing ergonomic assessments. We will use RULA (Rapid Upper Limb Assessment) and ROSA (Rapid Office Strain Assessment) for our project. Next, after outputting a value for each assessment, we can find out how much time surgeons spend in dangerous zones, and do a statistical analysis on our data. We also hope to include in the output ways that a surgeon could improve upon his or her posture, for example by decreasing neck flexion or adding support for their wrist.

Dependencies

We have a couple of dependencies, which will be enumerated and discussed.

1. C++ SDK, RULA/ROSA algorithms- understanding of these tools is crucial to our deliverables and dependencies, so some time will be spent getting familiar with these tools and assessments.

2. Images and videos for test data- if test data is not provided, it will be difficult for us to meet our deliverables as well because we cannot really assess the accuracy of our program, so we may create our own test data for this.

3. Wrist orientation, which is usually hidden from cameras but is used in RULA- we will edit the existing algorithm with an initial input asking if the surgeon's wrist is supported or not. This may change the output slightly.

4. Positioning and number of cameras- some videos at different angles may output different scores, so we need to test whether the significant enough to worry about and whether we need multiple angles for our input data.

Milestones and Status

  1. Milestone name: Single frame RULA/ROSA scores
    • Planned Date: March 4
    • Expected Date: March 25
    • Status: Delayed
  2. Milestone name: Video RULA/ROSA scores, User interface
    • Planned Date: March 16
    • Expected Date: April 15
    • Status: N/A
  3. Milestone name: Statistical Analyses, Decide # inputs
    • Planned Date: April 15
    • Expected Date: May 6
    • Status: N/A

Reports and presentations

Project Bibliography

Jóźwiak Z, Makowiec Dąbrowska T, Gadzicka E, Siedlecka J, Szyjkowska A, Kosobudzki M, Viebig P, Bortkiewicz A. Zastosowanie metody ROSA do oceny obciążenia układu mięśniowo-szkieletowego na komputerowych stanowiskach pracy [Using of the ROSA method to assess the musculoskeletal load on computer workstations]. Med Pr. 2019 Dec 3;70(6):675-699. Polish. doi: 10.13075

Davudian-Talab, Amirhossein & Azari, Gholamreza & Badfar, Gholamreza & Shafeei, Asrin & Derakhshan, Zainab. (2017). Evaluation and Correlation of the Rapid Upper Limb Assessment and Rapid Office Strain Assessment Methods for Predicting the Risk of Musculoskeletal Disorders. Internal Medicine and Medical Investigation Journal. 2. 155. 10.24200.

Namwongsa S, Puntumetakul R, Neubert MS, Chaiklieng S, Boucaut R (2018) Ergonomic risk assessment of smartphone users using the Rapid Upper Limb Assessment (RULA) tool. PLoS ONE 13(8): e0203394.

Rodrigues, Mirela Sant’Ana et al. ‘Differences in Ergonomic and Workstation Factors Between Computer Office Workers with and Without Reported Musculoskeletal Pain’. 1 Jan. 2017.

AIP Conference Proceedings 1883, 020034 (2017); https://doi.org/10.1063/1.5002052 Published Online: 14 September 2017

Dabholkar, Twinkle Yogesh et al. “Objective ergonomic risk assessment of wrist and spine with motion analysis technique during simulated laparoscopic cholecystectomy in experienced and novice surgeons.” Journal of minimal access surgery vol. 13,2 (2017): 124-130. doi:10.4103/0972-9941.195574

Gómez-Galán, M.; Callejón-Ferre, Á.-J.; Pérez-Alonso, J.; Díaz-Pérez, M.; Carrillo-Castrillo, J.-A. Musculoskeletal Risks: RULA Bibliometric Review. Int. J. Environ. Res. Public Health 2020, 17, 4354. https://doi.org/10.3390/ijerph17124354

Other Resources and Project Files

Here give list of other project files (e.g., source code) associated with the project. If these are online give a link to an appropriate external repository or to uploaded media files under this name space (2021-17).