What question did this study set out to answer?

Establish an advanced benchmark dataset to evaluate computer vision models in segmenting surgical instruments, medical devices, and anatomical structures during robotic surgery.

January 30, 2020Open Access

2018 Robotic Scene Segmentation Challenge

View Full Paper

Ask AI

Bookmark

Establish an advanced benchmark dataset to evaluate computer vision models in segmenting surgical instruments, medical devices, and anatomical structures during robotic surgery.
Acquired endoscopic surgical video using da Vinci robotic instruments operating on porcine tissue backgrounds with realistic motion dynamics.
Expanded segmentation classes from binary and articulating instrument parts to include multiple anatomical objects and medical devices.
Evaluated participant submissions that leveraged convolutional neural networks, primarily modified U-Net architectures, across iterative challenge editions (2015, 2017, and 2018).
Transition from synthetic kinematic annotations on ex-vivo tissue (2015) to realistic porcine backgrounds (2017–2018) successfully tested deep learning robustness against complex surgical scenes.
Participating convolutional neural network architectures adapted effectively to multi-class scene segmentation encompassing both articulating robotic instruments and underlying organs.

Discussion

No takes yet. Share an insight, caveat, or question.

Implication

Benchmark challenge report demonstrates multi-class segmentation of robotic tools and anatomy in porcine models, highlighting advances in computer vision for robotic surgery.

Cite This Study

Allan et al. (2020) studied this question.

synapsesocial.com/papers/6a7ec67a5f7e1a0066a6cbfb https://doi.org/10.48550/arxiv.2001.11190

View Full Paper

Ask AI

Bookmark

Implication

Benchmark challenge report demonstrates multi-class segmentation of robotic tools and anatomy in porcine models, highlighting advances in computer vision for robotic surgery.

2018 Robotic Scene Segmentation Challenge

Key Points

Discussion

Implication

Cite This Study

2018 Robotic Scene Segmentation Challenge

Key Points

Discussion

Implication

Cite This Study

Also Consider

Also Consider