DeepSplit: Segmentation of Microscopy Images Using Multi-task Convolutional Networks

Andrew Torr, Doga Basaran, Julia Sero, Jens Rittscher, Heba Sailem

Research output: Chapter in Book/Report/Conference proceedingChapter in a published conference proceeding


Accurate segmentation of cellular structures is critical for automating the analysis of microscopy data. Advances in deep learning have facilitated extensive improvements in semantic image segmentation. In particular, U-Net, a model specifically developed for biomedical image data, performs multi-instance segmentation through pixel-based classification. However, approaches based on U-Net tend to merge touching cells in dense cell cultures, resulting in under-segmentation. To address this issue, we propose DeepSplit; a multi-task convolutional neural network architecture where one encoding path splits into two decoding branches. DeepSplit first learns segmentation masks, then explicitly learns the more challenging cell-cell contact regions. We test our approach on a challenging dataset of cells that are highly variable in terms of shape and intensity. DeepSplit achieves 90% cell detection coefficient and 90% Dice Similarity Coefficient (DSC) which is a significant improvement on the state-of-the-art U-Net that scored 70% and 84% respectively.

Original languageEnglish
Title of host publicationMedical Image Understanding and Analysis - 24th Annual Conference, MIUA 2020, Proceedings
EditorsBartlomiej W. Papiez, Ana I.L. Namburete, Mohammad Yaqub, J. Alison Noble, Mohammad Yaqub
Place of PublicationCham, Switzerland
Number of pages13
ISBN (Print)9783030527907
Publication statusE-pub ahead of print - 8 Jul 2020
Event24th Annual Conference on Medical Image Understanding and Analysis, MIUA 2020 - Oxford, UK United Kingdom
Duration: 15 Jul 202017 Jul 2020

Publication series

NameCommunications in Computer and Information Science
Volume1248 CCIS
ISSN (Print)1865-0929
ISSN (Electronic)1865-0937


Conference24th Annual Conference on Medical Image Understanding and Analysis, MIUA 2020
Country/TerritoryUK United Kingdom

ASJC Scopus subject areas

  • Computer Science(all)
  • Mathematics(all)


Dive into the research topics of 'DeepSplit: Segmentation of Microscopy Images Using Multi-task Convolutional Networks'. Together they form a unique fingerprint.

Cite this