Abstract
Abstract Surgical tool recognition is a key task to analyze surgical workflow, in order to improve the efficiency and safety of laparoscopic surgeries. The laparoscopic videos are important sources to conduct this task, However, there are some challenges to analyze these videos. Focus on the imbalanced dataset problem, data augmentation method based on generate different synthetic datasets and evaluate their performance training on a convolutional neural network model are investigated in this research. The results show the effect on the model with different background patterns. A better performance was achieved when the model was trained by a structure background dataset. Further research will be needed to understand why the original background patterns support the correct classification. It is assumed that this is an overlearning effect, that will not hold if other procedures were included into the test set.
Highlights
Laparoscopic videos contain valuable information of minimally invasive surgeries, such as surgical tools, surgical actions and tissues [1]
Surgical tool recognition based on the analysis of these laparoscopic videos have gained increasing attention by researchers due to its importance to better control surgical workflow
The final fully-connected layer is replaced to adjust to the cholecystectomy data, which shows seven different classes i.e. seven surgical tools are to be recognized
Summary
Laparoscopic videos contain valuable information of minimally invasive surgeries, such as surgical tools, surgical actions and tissues [1]. Some tools appear more frequently than others, causing the imbalanced dataset problem [1,2] It greatly affects the training process of CNN model. We take a first step toward generating synthetic data that can be used to augment available datasets in order to improve tool presence detection using CNNs. The influence of the background to the training efficiency is explored. Three artificial datasets were generated by substituting the image background by three different patterns, one structured and two unstructured backgrounds, originalbackgrounds, uniform-backgrounds and random-backgrounds. An evaluation of these datasets in terms of their effectivity to train a CNN model for tool detection was made
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
