The ITEC LapGyn4 Gynecologic Laparoscopy Image Dataset (v1.2)

The dataset comprises four individual datasets taken from 500+ gynecologic laparoscopic surgeries for the task of automatic content analysis. The individual collections contain image classes depicting general surgical actions, anatomical structures, conducted actions on specific anatomy as well as examples of differing amounts of visible instruments:

Surgical Actions

Surgical Actions are general activities performed during surgery involving one or more instruments.

Anatomical Structures

Anatomical Structures contains a collection of different inner bodily organs that are of significance for several laparoscopic procedures.

Actions on Anatomy

Actions on Anatomy comprises specific actions on particular anatomy and it at this point includes suturing on different organs.

Instrument Count *

The Instrument Count dataset contains images of different amounts of visible surgical tools.

Note: Not all appearing body-external objects are considered instruments. Within this dataset, following items are not counted:

Dataset details (sample counts)

Surgical Actions Anatomical Structures Actions on Anatomy Instrument Count
Suction & Irrigation 3036 Uterus 938 Suturing (Uterus) 940 0 Instr. 5104
Suturing 12914 Ovary 1162 Suturing (Ovary) 715 1 Instr. 5124
Cutting (C) 1185 Oviduct 195 Suturing (Vagina) 1022 2 Instr. 5734
Cutting (HF) 3752 Liver 138 Suturing (Other) 2110 3 Instr. 5462
Dissection (blunt) 1444 Colon 295
Sling (Hyst.) 2752
Coagulat. 3480
Injection 2119
Total 30682 2728 4787 21424

For further information about the dataset's organization see 'Readme.txt' within the download archive below.

Version History

All previous dataset revisions are listed here and their release pages are linked.

v1.2 (this version)


Reference and Copyrights

The dataset is 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. Jörg Keckstein.

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

A. Leibetseder, S. Petscharnig, M. J. Primus, S. Kletz, B. Münzer, K. Schoeffmann, J. Keckstein, LapGyn4: A Dataset for 4 Automatic Content Analysis Problems in the Domain of Laparoscopic Gynecology, ACM Multimedia Systems Conference 2018

author    = {Andreas Leibetseder and
            Stefan Petscharnig and
            Manfred J{\"{u}}rgen Primus and
            Sabrina Kletz and
            Bernd M{\"{u}}nzer and
            Klaus Schoeffmann and
            J{\"{o}}rg Keckstein},
title     = {Lapgyn4: a dataset for 4 automatic content analysis problems in the
            domain of laparoscopic gynecology},
booktitle = {Proceedings of the 9th {ACM} Multimedia Systems Conference, MMSys
            2018, Amsterdam, The Netherlands, June 12-15, 2018},
pages     = {357--362},
publisher = {{ACM}},
year      = {2018},
url       = {},
doi       = {10.1145/3204949.3208127},
timestamp = {Wed, 21 Nov 2018 12:44:03 +0100},
biburl    = {},
bibsource = {dblp computer science bibliography,}

* Certain images of the Instrument Count dataset are extracted from the Cholec80 dataset, hence, when utilizing this dataset you are requested to as well refer to Twinada et al. [1].

LapGyn4 is licensed under Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0, Creative Commons License) and is created as well as maintained by Distributed Multimedia Systems Group of the Institute of Information Technology (ITEC) at Alpen-Adria Universität 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 LapGyn4_v1.2 (~11GB).
The mapping files for the exact our exact 5-fold cross validation split can be downloaded on v1.0 page, but note that file names and dataset size in case of 'Actions on Anatomy' have changed, therefore, they no longer can be applied to this dataset version (v1.2).

[1] A.P. Twinanda, S. Shehata, D. Mutter, J. Marescaux, M. de Mathelin, N. Padoy, EndoNet: A Deep Architecture for Recognition Tasks on Laparoscopic Videos, IEEE Transactions on Medical Imaging (TMI), to appear (arXiv preprint), doi:10.1109/TMI.2016.2593957, 2016.