Skip to content
PhaKIR - Challenge:
Phase, Keypoint and Instrument Recognition

part of

PhaKIR - Challenge:
Phase, Keypoint and Instrument Recognition

part of

PhaKIR - Challenge:
Phase, Keypoint and Instrument Recognition

part of

ABOUT THE CHALLENGE

Accurate and reliable recognition and localization of surgical instruments in endoscopic video recordings is the basis for a variety of applications in computer- and robot-assisted minimally invasive surgery (RAMIS) [1]. The robust handling of real-world conditions such as varying illumination levels, blurred movement of the instruments and the camera, severe sudden bleeding that impairs the field of view, or even unexpected smoke development is an important prerequisite for such procedures. To process the information extracted from the endoscopic images in the best possible way, the inclusion of the context of the operation can be used as a promising possibility, which can be realized, for example, by knowing the current phase of an intervention.

In our EndoVis2024 subchallenge, we present a dataset for which three tasks are to be performed: Instance segmentation of the surgical instruments, keypoint estimation, and procedure phase recognition. The following annotations are available for this: pixel-accurate instance segmentations of surgical instruments together with their instrument types for a total of 19 categories, coordinates of relevant instrument keypoints (instrument tip(s), shaft-tip transition, shaft), and a classification of the phases of an intervention into eight different phase categories. Our dataset consists of 13 real-world videos of human cholecystectomies ranging from 23 to 60 minutes in duration. The procedures were performed by experienced physicians, and the videos were recorded in three hospitals. In addition to existing datasets, our annotations provide instance segmentations of surgical instruments, relevant keypoints, and intervention phases in one dataset and thus comprehensively cover instrument localization and the context of the operation. Furthermore, the provision of the complete video sequences offers the opportunity to include the temporal information regarding the respective tasks and thus further optimize the resulting methods and outcomes.

[1] T. Rueckert, D. Rueckert, and C. Palm, "Methods and datasets for segmentation of minimally invasive surgical instruments in endoscopic images and videos: A review of the state of the art", Computers in Biology and Medicine, vol. 169, pp. 107929, 2024, DOI: https://doi.org/10.1016/j.compbiomed.2024.107929.

 

AWARDS

For each of the three tasks:

1st place with 500 €
2nd place with 300 €
3rd place with 200 €


Sponsored by

ABOUT THE CHALLENGE

Accurate and reliable recognition and localization of surgical instruments in endoscopic video recordings is the basis for a variety of applications in computer- and robot-assisted minimally invasive surgery (RAMIS) [1]. The robust handling of real-world conditions such as varying illumination levels, blurred movement of the instruments and the camera, severe sudden bleeding that impairs the field of view, or even unexpected smoke development is an important prerequisite for such procedures. To process the information extracted from the endoscopic images in the best possible way, the inclusion of the context of the operation can be used as a promising possibility, which can be realized, for example, by knowing the current phase of an intervention.

In our EndoVis2024 subchallenge, we present a dataset for which three tasks are to be performed: Instance segmentation of the surgical instruments, keypoint estimation, and procedure phase recognition. The following annotations are available for this: pixel-accurate instance segmentations of surgical instruments together with their instrument types for a total of 19 categories, coordinates of relevant instrument keypoints (instrument tip(s), shaft-tip transition, shaft), and a classification of the phases of an intervention into eight different phase categories. Our dataset consists of 13 real-world videos of human cholecystectomies ranging from 23 to 60 minutes in duration. The procedures were performed by experienced physicians, and the videos were recorded in three hospitals. In addition to existing datasets, our annotations provide instance segmentations of surgical instruments, relevant keypoints, and intervention phases in one dataset and thus comprehensively cover instrument localization and the context of the operation. Furthermore, the provision of the complete video sequences offers the opportunity to include the temporal information regarding the respective tasks and thus further optimize the resulting methods and outcomes.

[1] T. Rueckert, D. Rueckert, and C. Palm, "Methods and datasets for segmentation of minimally invasive surgical instruments in endoscopic images and videos: A review of the state of the art", Computers in Biology and Medicine, vol. 169, pp. 107929, 2024, DOI: https://doi.org/10.1016/j.compbiomed.2024.107929.

 

AWARDS

For each of the three tasks:

1st place with 500 €
2nd place with 300 €
3rd place with 200 €


Sponsored by

TIMELINE

April 2024 Challenge Registration Opens
June 2024 Release of  1st Part of Training Data
July 2024 Release of 2nd Part of Training Data
1st August 2024 Release of Docker Submission Guide and Evaluation Instructions
8th September 2024 Start of Docker Submissions to Verify Functionality
15th September 2024 Submission Deadline and Registration Closing
15h September 2024 Submission Deadline Methodology Report
Day of EndoVis 2024 Challenge Day and Presentation of Results

PARTICIPATION POLICIES

External Training Data: Only training data provided by the challenge organizers and publicly available data sets are allowed for training. In addition, networks that have been pre-trained on publicly available datasets may be used.

Award Policy: We will name the 1st, 2nd, and 3rd place for each task separately. We provide certificates and award money (500€, 300€, 200€) for all of them, sponsored by AKTORmed GmbH.

Publication Policy: All members of the participating teams are qualified as authors. All participants are allowed to publish their own results separately after the challenge but need to cite the final challenge publication created and published by the organizers of the challenge. Individual publications have an embargo time and may only be published once the initial challenge publication from the organizers has been uploaded to arXiv or another platform.

Licensing: The dataset is published under a Creative Commons Attribution-NonCommercial-ShareAlike (CC BY-NC-SA) license, which means that it can be used for non-commercial purposes once the challenge has been conducted and the challenge paper has been published. If you wish to use or reference this dataset, you must cite this challenge paper that will appear after the challenge. The licensing of new creations must use the exact same licensing terms as in the current version of the dataset.

We use cookies to improve the user-friendliness of the website. By visiting our website you agree to this.