Dataset release for "Event Recognition in Laparoscopic Gynecology Videos with Hybrid Transformers"

The Laparoscopic Gynecology dataset comprises a collection of more than 600 surgical videos capturing laparoscopic procedures conducted at the Medical University of Vienna. Within this extensive dataset, we provide surgical event recognition for a subset of 174 surgical videos, along with annotations for four distinct relevant events. These events include Abdominal access, Bleeding, Coagulation and Transection, and Needle passing. To provide a clearer understanding, we offer examples of each of these surgical events within the dataset.

We provide a large annotated dataset to enable comprehensive studies on deep-learning-based event recognition in laparoscopic gynecology surgery videos. Table 1 presents a visual representation of annotations for selected videos from our laparoscopic surgery dataset. Each row in this table corresponds to a unique laparoscopic surgery case, with durations varying from 11 minutes to approximately three hours. Within every individual case, the presence of distinct events, including Abdominal Access, Bleeding, Coalulation/Transection, and Needle Passing, along with irrelevant events, is depicted using different colors. The segments within these videos also vary in duration, ranging from as short as one second to over one minute in some instances. Furthermore, Table 2 provides an overview of the number of segments and the total duration of annotations associated with each specific event.

Within the Event Recognition LapGyn dataset, each event is organized into both training and testing sets, structured to facilitate binary classification tasks. The video clips in these sets typically have durations of 2 to 3 seconds. Additionally, our dataset extends to the Event Segments LapGyn dataset, which provides extracted segments from video cases. For each specific event, we provide two folders: one containing relevant segments and another containing irrelevant segments.


The datasets are exclusively provided for scientific research purposes and as such cannot be used commercially or for any other purpose. If any other purpose is intended, you may directly contact the originator of the videos, Prof. Dr. Heinrich Husslein or Assoc. Prof. DI Dr. Klaus Schoeffmann.

Besides, a reference must be made to the following publication when this dataset is used in any academic and research reports:

Nasirihaghighi, S., Ghamsarian, N., Husslein, H., Schoeffmann, K.: Event Recognition in Laparoscopic Gynecology Videos with Hybrid Transformers

@inproceedings{Event Recognition,
    author    = {Sahar Nasirihaghighi and
		Negin Ghamsarian and
                Heinrich Husslein and
                Klaus Schoeffmann},
    title     = {Event Recognition in Laparoscopic Gynecology Videos with Hybrid Transformers},

The datasets are licensed under Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0, Creative Commons License) and are created as well as maintained by Distributed Multimedia Systems Group of the Institute of Information Technology (ITEC) at Alpen-Adria University in Klagenfurt, Austria.

This license allows users of this dataset to copy, distribute, and transmit the work under the following conditions:

For further legal details, please read the complete license terms.


If you agree to above conditions, you are free to download: