Multi-label Classification of Surgical Tools with Convolutional Neural Networks

  • 2018-05-15 13:39:54
  • Jonas Prellberg, Oliver Kramer
  • 4


Automatic tool detection from surgical imagery has a multitude of usefulapplications, such as real-time computer assistance for the surgeon. Using thesuccessful residual network architecture, a system that can distinguish 21different tools in cataract surgery videos is created. The videos are providedas part of the 2017 CATARACTS challenge and pose difficulties found in manyreal-world datasets, for example a strong class imbalance. The construction ofthe detection system is guided by a wide array of experiments that exploredifferent design decisions.


Introduction (beta)



Conclusion (beta)