Video retrieval in laparoscopic video recordings with dynamic content descriptors

Abstract

In the domain of gynecologic surgery an increasing number of surgeries are performed in a minimally invasive manner. These laparoscopic surgeries require specific psychomotor skills of the operating surgeon, which are difficult to learn and teach. This is the reason why an increasing number of surgeons promote checking video recordings of laparoscopic surgeries for the occurrence of technical errors with surgical actions. This manual surgical quality assessment (SQA) process, however, is very cumbersome and timeconsuming when carried out without any support from content-based video retrieval. Appl (2018) 77:16813-16832 Descriptor) that can be effectively used to find similar segments in a laparoscopic video database and thereby help surgeons to more quickly inspect other instances of a given error scene. We evaluate the retrieval performance of MIDD with surgical actions from gynecologic surgery in direct comparison to several other dynamic content descriptors. We show that the MIDD descriptor significantly outperforms the state-of-the-art in terms of retrieval performance as well as in terms of runtime performance. Additionally, we release the manually created video dataset of 16 classes of surgical actions from medical laparoscopy to the public, for further evaluations.

wildNeRF: Novel view synthesis of in-the-wild dynamic scenes using using sparse monocular view data

“…In this paper, we show quantitative and qualitative results on the NVIDIA Dynamic Scenes Dataset [18], and qualitative results (due to the lack of ground truth depth data) on complex surgical datasets, Cholec80 [20] and SugicalAc-tions160 [19], which deal with significant unstructured motion between the camera and the objects in a scene.…”

“…Our contributions are as follows: 1. We showcase the efficacy of our model using the publicly available NVIDIA dynamic scenes dataset and extend our model to real-world datasets such as the SurgicalActions160 [19] and Cholec80 [20] datasets. 2.…”

A Review on Deep Learning in Minimally Invasive Surgery

“…• SurgicalActions160 12 : this dataset contains 10 examples of 16 classes of surgical actions (160 videos in total). These actions are subject to surgical errors so that they constitute good testing data in the context of surgical skill assessment [23]. • LapSig300: it is the first large-scale dataset of laparoscopic colorectal surgery.…”

Hand Motion-Aware Surgical Tool Localization and Classification from an Egocentric Camera

“…Studies of surgical tool detection have been mainly focused on procedures in which the surgeon sees the surgical field not directly but through a camera, such as cataract and endoscopic surgery. There are some datasets for the surgical tool detection, such as Cholec80 [ 23 ], ITEC [ 24 ], M2CAI2016 [ 18 ], CATARACT2017 [ 25 ] dataset. These datasets cannot be applied to our task as the domain of the surgery in these datasets is different from the domain of the open surgery.…”