Automated Classification of Apoptosis in Phase Contrast Microscopy Using Capsule Network

Automatic and accurate classification of apoptosis, or programmed cell death, will facilitate cell biology research. The state-of-the-art approaches in apoptosis classification use deep convolutional neural networks (CNNs). However, these networks are not efficient in encoding the part-whole relationships, thus requiring a large number of training samples to achieve robust generalization. This paper proposes an efficient variant of capsule networks (CapsNets) as an alternative to CNNs. Extensive experimental results demonstrate that the proposed CapsNets achieve competitive performances in target cell apoptosis classification, while significantly outperforming CNNs when the number of training samples is small. To utilize temporal information within microscopy videos, we propose a recurrent CapsNet constructed by stacking a CapsNet and a bi-directional long short-term recurrent structure. Our experiments show that when considering temporal constraints, the recurrent CapsNet achieves 93.8% accuracy and makes significantly more consistent prediction than NNs.

[1]  C. Reutelingsperger,et al.  Annexin V-affinity assay: a review on an apoptosis detection system based on phosphatidylserine exposure. , 1998, Cytometry.

[2]  Takeo Kanade,et al.  Understanding the phase contrast optics to restore artifact-free microscopy images for segmentation , 2012, Medical Image Anal..

[3]  J Cummings,et al.  Biomarkers of apoptosis , 2008, British Journal of Cancer.

[4]  Harjeet Singh,et al.  Automated profiling of individual cell-cell interactions from high-throughput time-lapse imaging microscopy in nanowell grids (TIMING) , 2015, Bioinform..

[5]  Stephan J. Garbin,et al.  Harmonic Networks: Deep Translation and Rotation Equivariance , 2016, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[6]  Hien Van Nguyen,et al.  Fast CapsNet for Lung Cancer Screening , 2018, MICCAI.

[7]  Jian Sun,et al.  Deep Residual Learning for Image Recognition , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[8]  S. Eddy.  Hidden Markov models. , 1996, Current opinion in structural biology.

[9]  J. Ó. Neill.  Siamese Capsule Networks , 2018, 1805.07242.

[10]  Androulla N. Miliotou,et al.  CAR T-cell Therapy: A New Era in Cancer Immunotherapy. , 2018, Current pharmaceutical biotechnology.

[11]  Kai Sundmacher,et al.  A guide to automated apoptosis detection: How to make sense of imaging flow cytometry data , 2018, PloS one.

[12]  Jürgen Schmidhuber,et al.  Long Short-Term Memory , 1997, Neural Computation.

[13]  Miroslav Hejna,et al.  High accuracy label-free classification of single-cell kinetic states from holographic cytometry of human melanoma cells , 2017, Scientific Reports.

[14]  Kilian Q. Weinberger,et al.  Densely Connected Convolutional Networks , 2016, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[15]  Max Welling,et al.  Group Equivariant Convolutional Networks , 2016, ICML.

[16]  Premkumar Natarajan,et al.  CapsuleGAN: Generative Adversarial Capsule Network , 2018, ECCV Workshops.

[17]  Mohammad Taha Bahadori,et al.  Spectral Capsule Networks , 2018 .

[18]  Geoffrey E. Hinton,et al.  ATRIX CAPSULES WITH EM ROUTING , 2018 .

[19]  S. Elmore.  Apoptosis: A Review of Programmed Cell Death , 2007, Toxicologic pathology.

[20]  Donald Wlodkowic,et al.  Dynamic analysis of apoptosis using cyanine SYTO probes: from classical to microfluidic cytometry. , 2009, Experimental cell research.

[21]  Chen Xu,et al.  MS-CapsNet: A Novel Multi-Scale Capsule Network , 2018, IEEE Signal Processing Letters.

[22]  H. Steller,et al.  Programmed Cell Death in Animal Development and Disease , 2011, Cell.

[23]  Todd Thorsen,et al.  High-density microfluidic arrays for cell cytotoxicity analysis. , 2007, Lab on a chip.

[24]  S. Kain,et al.  Early detection of apoptosis using a fluorescent conjugate of annexin V. , 1997, BioTechniques.

[25]  G. Gores,et al.  A novel cytotoxicity screening assay using a multiwell fluorescence scanner. , 1992, Toxicology and applied pharmacology.

[26]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[27]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[28]  C Haanen,et al.  A novel assay for apoptosis. Flow cytometric detection of phosphatidylserine expression on early apoptotic cells using fluorescein labelled Annexin V. , 1995, Journal of immunological methods.

[29]  Harjeet Singh,et al.  TIMING 2.0: high‐throughput single‐cell profiling of dynamic cell‐cell interactions by time‐lapse imaging microscopy in nanowell grids , 2018, Bioinform..

[30]  Geoffrey E. Hinton,et al.  Dynamic Routing Between Capsules , 2017, NIPS.

[31]  Anne E Carpenter,et al.  Reconstructing cell cycle and disease progression using deep learning , 2017, Nature Communications.

[32]  Brian T. Cunningham,et al.  Label-free imaging of cancer cells using photonic crystal biosensors and application to cytotoxicity screening of a natural compound library , 2008 .

[33]  Michel Sadelain,et al.  Chimeric Antigen Receptor Therapy. , 2018, The New England journal of medicine.

[34]  Takeo Kanade,et al.  Apoptosis Detection for Adherent Cell Populations in Time-Lapse Phase-Contrast Microscopy Images , 2012, MICCAI.

[35]  Ulas Bagci,et al.  Capsules for Object Segmentation , 2018, ArXiv.

[36]  Jérôme Parent,et al.  Label-free cytotoxicity screening assay by digital holographic microscopy. , 2013, Assay and drug development technologies.

[37]  Andrew McCallum,et al.  Conditional Random Fields: Probabilistic Models for Segmenting and Labeling Sequence Data , 2001, ICML.

[38]  Gideon Kowadlo,et al.  Sparse Unsupervised Capsules Generalize Better , 2018, ArXiv.