Sparse Data-Driven Learning for Effective and Efficient Biomedical Image Segmentation

Annu Rev Biomed Eng. 2020 Jun 4:22:127-153. doi: 10.1146/annurev-bioeng-060418-052147. Epub 2020 Mar 13.

Abstract

Sparsity is a powerful concept to exploit for high-dimensional machine learning and associated representational and computational efficiency. Sparsity is well suited for medical image segmentation. We present a selection of techniques that incorporate sparsity, including strategies based on dictionary learning and deep learning, that are aimed at medical image segmentation and related quantification.

Keywords: dictionary learning; image representation; image segmentation; machine learning; medical image analysis; sparsity.

Publication types

  • Research Support, N.I.H., Extramural
  • Review

MeSH terms

  • Algorithms
  • Animals
  • Brain / diagnostic imaging
  • Deep Learning
  • Dogs
  • Echocardiography / methods
  • Heart Ventricles / diagnostic imaging
  • Humans
  • Image Processing, Computer-Assisted / methods*
  • Imaging, Three-Dimensional / methods*
  • Machine Learning
  • Models, Theoretical
  • Neural Networks, Computer
  • Tomography, X-Ray Computed / methods