Learning to detect cells using non-overlapping extremal regions

Conference item

Cell detection in microscopy images is an important step in the automation of cell based-experiments. We propose a machine learning-based cell detection method applicable to different modalities. The method consists of three steps: first, a set of candidate cell-like regions is identified. Then, each candidate region is evaluated using a statistical model of the cell appearance. Finally, dynamic programming picks a set of non-overlapping regions that match the model. The cell model requires few images with simple dot annotation for training and can be learned within a structured SVM framework. In the reported experiments, state-of-the-art cell detection accuracy is achieved for H&E stained histology, fluorescence, and phase-contrast images.

Authors: Arteta, C; Lempitsky, V; Noble, JA; Zisserman, A

• Publication status: Published
• Peer review status: Peer reviewed
• Publisher: Springer
• Host title: Medical Image Computing and Computer-Assisted Intervention -- MICCAI 2012 15th International Conference, Nice, France, October 1-5, 2012, Proceedings, Part I
• Pages: 348-356
• Series: Lecture Notes in Computer Science
• Series number: 7510
• Publication date: 2012-09-21
• Event title: 15th International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI 2012)
• Event location: Nice, France
• Event start date: 2012-10-01
• Event end date: 2012-10-05
• DOI: 10.1007/978-3-642-33415-3_43
• EISSN: 1611-3349
• ISSN: 0302-9743
• Pmid: 23285570
• EISBN: 978-3-642-33415-3
• ISBN: 978-3-642-33414-6
• Language: English
• Keywords: phase-contrast; algorithms; microscopy; cell size; automated; statistical; models; reproducibility of results; hela cells; computer-assisted; pattern recognition; software; computer simulation; image processing; artificial intelligence; support vector machine