Open AccessA Comparative Study of Generic Visual Components of Two‐Dimensional Versus Three‐Dimensional Laparoscopic Images
Author(s) -
El Boghdady Michael,
Ramakrishnan Gobinath,
Tang Benjie,
Alijani Afshin
Publication year - 2018
Publication title -
world journal of surgery
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.115
H-Index - 148
eISSN - 1432-2323
pISSN - 0364-2313
DOI - 10.1007/s00268-017-4220-3
Subject(s) - artificial intelligence , computer vision , object (grammar) , curvature , computer science , spatial reference system , reliability (semiconductor) , mathematics , geometry , physics , power (physics) , quantum mechanics
Aims There is a strong evidence to suggest that 3D imaging improves the laparoscopic task performance when compared against 2D. However, to date, no study has explained why that might be. We identified six generic visual components during laparoscopic imaging and aimed to study each component in both 2D and 3D environments for comparison. Methods Twenty‐four consented laparoscopic novices performed specific isolated tasks in a laparoscopic Endo Trainer in 2D and 3D separately. The six endpoints were the accuracy in detecting changes in the laparoscopic images in the following components: distance, area, angle, curvature, volume and spatial coordinates. All the components except the spatial coordinates were assessed by creation, measurement and comparison. Each component was analysed between 2D and 3D groups and within each group at different values. Tests of spatial coordinates were video‐recorded and analysed for error number and error types by human reliability analysis technique. Errors types included past‐pointing, not reaching the object and touching the wrong object. The results were statistically analysed with independent T test. Results There was no statistically significant difference between 2D and 3D accuracy in the angle, area, distance and curvature. 3D performed more accurately in comparing volumes ( p = 0.05). In spatial coordinates, there were a statistically significant higher number of errors in 2D as compared to 3D ( p < 0.001). Past‐pointing and touching the wrong objects were significantly higher in 2D ( p < 0.05). Conclusion Between all the visual components, detecting change in volume and the spatial coordinates showed significant improvement in 3D environment when compared to 2D.