A machine learning approach for real-time modelling of tissue deformation in image-guided neurosurgery

作者：

Highlights：

• A real-time soft tissue deformation computation method is presented.

• The method uses machine learning to derive models from the results of multiple load-driven FEM simulations.

• A brain tumour is used as the subject of the deformation model.

• Once trained, the models can predict the deformation of the tumour in real-time with relative positional errors below 0.3 mm.

摘要

•A real-time soft tissue deformation computation method is presented.•The method uses machine learning to derive models from the results of multiple load-driven FEM simulations.•A brain tumour is used as the subject of the deformation model.•Once trained, the models can predict the deformation of the tumour in real-time with relative positional errors below 0.3 mm.

论文关键词：Soft tissue deformation,Biomechanics,Finite element method,Image-guided surgery,Machine learning,Artificial neural networks,Support vector regression

论文评审过程：Received 3 November 2016, Revised 19 May 2017, Accepted 6 July 2017, Available online 24 July 2017, Version of Record 7 September 2017.

论文官网地址：https://doi.org/10.1016/j.artmed.2017.07.004