Classification of Craniosynostosis on 2D Asymmetry Maps Using Convolutional Neural Networks

  • type:Bachelor thesis
  • tutor:

    M.Sc. Matthias Schaufelberger

  • person in charge:

    Bahadur Savaş

  • Craniosynostosis is characterized by the premature fusion of sutures in an infant skull. By projecting an asymmetry metric from the vertex positions of a 3D surface mesh onto a 2D plane, asymmetry maps can be created. We have the hypothesis that such an asymmetry map can be used to classify craniosynostosis using Convolutional Neural Networks (CNNs).