Surgical phase and instrument recognition: how to identify appropriate dataset splits.

Journal: International journal of computer assisted radiology and surgery
Published Date:

Abstract

PURPOSE: Machine learning approaches can only be reliably evaluated if training, validation, and test data splits are representative and not affected by the absence of classes. Surgical workflow and instrument recognition are two tasks that are complicated in this manner, because of heavy data imbalances resulting from different length of phases and their potential erratic occurrences. Furthermore, sub-properties like instrument (co-)occurrence are usually not particularly considered when defining the split.

Authors

  • Georgii Kostiuchik
    Department of Cardiac Surgery, Heidelberg University Hospital, Heidelberg, Germany. georgii.kostiuchik@med.uni-heidelberg.de.
  • Lalith Sharan
    Department of Cardiac Surgery, Heidelberg University Hospital, Heidelberg, Germany.
  • Benedikt Mayer
    Department of Simulation and Graphics, University of Magdeburg, Magdeburg, Germany.
  • Ivo Wolf
  • Bernhard Preim
    Department of Simulation and Graphics, Otto von Guericke University Magdeburg, Germany; Research Campus STIMULATE, Otto von Guericke University Magdeburg, Germany.
  • Sandy Engelhardt