Prangemeier, Tim ; Reich, C. ; Koeppl, H. (2020):
Attention-Based Transformers for Instance Segmentation of Cells in Microstructures.
In: Proceedings: 2020 IEEE International Conference on Bioinformatics and Biomedicine,
IEEE, IEEE International Conference on Bioinformatics and Biomedicine (IEEE BIBM 2020), virtual Conference, 16.-19.12.2020, ISBN 978-1-7281-6215-7,
DOI: 10.1109/BIBM49941.2020.9313305,
[Conference or Workshop Item]

Abstract

Detecting and segmenting object instances is acommon task in biomedical applications. Examples range from detecting lesions on functional magnetic resonance images, to the detection of tumours in histopathological images and extracting quantitative single-cell information from microscopy imagery, where cell segmentation is a major bottleneck. Attention-basedtransformers are state-of-the-art in a range of deep learningfields. They have recently been proposed for segmentation taskswhere they are beginning to outperforming other methods. We present a novel attention-basedcell detection transformer(Cell-DETR) for direct end-to-end instance segmentation. While the segmentation performance is on par with a state-of-the-artinstance segmentation method, Cell-DETR is simpler and faster.We showcase the method’s contribution in a the typical use caseof segmenting yeast in microstructured environments, commonly employed in systems or synthetic biology. For the specific usecase, the proposed method surpasses the state-of-the-art tools forsemantic segmentation and additionally predicts the individual object instances. The fast and accurate instance segmentation performance increases the experimental information yield for a posterioridata processing and makes online monitoring of experiments and closed-loop optimal experimental design feasible.

Send an inquiry

Show editorial Details