Few-shot hypercolumn-based mitochondria segmentation in cardiac and outer hair cells in focused ion beam-scanning electron microscopy (FIB-SEM) data

Julia Dietlmeier, Kevin McGuinness, Sandra Rugonyi, Teresa Wilson, Alfred Nuttall, Noel E. O'Connor

Research output: Contribution to journalArticle


We present a novel AI-based approach to the few-shot automated segmentation of mitochondria in large-scale electron microscopy images. Our framework leverages convolutional features from a pre-trained deep multilayer convolutional neural network, such as VGG-16. We then train a binary gradient boosting classifier on the resulting high-dimensional feature hypercolumns. We extract VGG-16 features from the first four convolutional blocks and apply bilinear upsampling to resize the obtained maps to the input image size. This procedure yields a 2688-dimensional feature hypercolumn for each pixel in a 224 × 224 input image. We then apply L1-regularized logistic regression for supervised active feature selection to reduce dependencies among the features, to reduce overfitting, as well as to speed-up gradient boosting-based training. During inference we block process 1728 × 2022 large microscopy images. Our experiments show that in such a formulation of transfer learning our processing pipeline is able to achieve high-accuracy results on very challenging datasets containing a large number of irregularly shaped mitochondria in cardiac and outer hair cells. Our proposed few-shot training approach gives competitive performance with the state-of-the-art using far less training data.

Original languageEnglish (US)
Pages (from-to)521-528
Number of pages8
JournalPattern Recognition Letters
StatePublished - Dec 1 2019



  • Cardiac and outer hair cells
  • Deep learning
  • Few-shot learning
  • Gradient boosting
  • Mitochondria segmentation
  • Transfer learning

ASJC Scopus subject areas

  • Software
  • Signal Processing
  • Computer Vision and Pattern Recognition
  • Artificial Intelligence

Cite this