Iterative Deep Learning Based Unbiased Stereology with Human-in-the-Loop

Lack of enough labeled data is a major problem in building machine learning based models when the manual annotation (labeling) is error-prone, expensive, tedious, and time-consuming. In this paper, we introduce an iterative deep learning based method to improve segmentation and counting of cells based on unbiased stereology applied to regions of interest of extended depth of field (EDF) images. This method uses an existing machine learning algorithm called the adaptive segmentation algorithm (ASA) to generate masks (verified by a user) for EDF images to train deep learning models. Then an iterative deep learning approach is used to feed newly predicted and accepted deep learning masks/images (verified by a user) to the training set of the deep learning model. The error rate in unbiased stereology count of cells on an unseen test set reduced from about 3 % to less than 1 % after 5 iterations of the iterative deep learning based unbiased stereology process.

[1]  Guigang Zhang,et al.  Deep Learning , 2016, Int. J. Semantic Comput..

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

[3]  Nitish Srivastava,et al.  Dropout: a simple way to prevent neural networks from overfitting , 2014, J. Mach. Learn. Res..

[4]  Peter R. Mouton,et al.  Unbiased Stereology: A Concise Guide , 2011 .

[5]  Patrice Y. Simard,et al.  Best practices for convolutional neural networks applied to visual document analysis , 2003, Seventh International Conference on Document Analysis and Recognition, 2003. Proceedings..

[6]  H. Gundersen,et al.  Unbiased stereological estimation of the total number of neurons in the subdivisions of the rat hippocampus using the optical fractionator , 1991, The Anatomical record.

[7]  A G Valdecasas,et al.  On the extended depth of focus algorithms for bright field microscopy. , 2001, Micron.

[8]  B. Hyman,et al.  Region-specific dissociation of neuronal loss and neurofibrillary pathology in a mouse model of tauopathy. , 2006, The American journal of pathology.

[9]  Bram van Ginneken,et al.  A survey on deep learning in medical image analysis , 2017, Medical Image Anal..

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

[11]  Geoffrey E. Hinton,et al.  Rectified Linear Units Improve Restricted Boltzmann Machines , 2010, ICML.

[12]  A. Bradley,et al.  A One-Pass Extended Depth of Field Algorithm Based on the Over-Complete Discrete Wavelet Transform , 2004 .

[13]  A. Savitzky,et al.  Smoothing and Differentiation of Data by Simplified Least Squares Procedures. , 1964 .

[14]  Lawrence O. Hall,et al.  Unbiased estimation of cell number using the automatic optical fractionator , 2017, Journal of Chemical Neuroanatomy.

[15]  Geoffrey E. Hinton,et al.  Reducing the Dimensionality of Data with Neural Networks , 2006, Science.

[16]  Alexandra Flemming,et al.  Infectious disease: Unravelling SARS lethality , 2005, Nature Reviews Drug Discovery.

[17]  Thomas Brox,et al.  U-Net: Convolutional Networks for Biomedical Image Segmentation , 2015, MICCAI.

[18]  Sergey Ioffe,et al.  Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift , 2015, ICML.

[19]  Maurice Ptito,et al.  Knowing What Counts: Unbiased Stereology in the Non-human Primate Brain , 2009, Journal of visualized experiments : JoVE.

[20]  Martín Abadi,et al.  TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems , 2016, ArXiv.

[21]  C. Saper,et al.  Any way you cut it: A new journal policy for the use of unbiased counting methods , 1996, The Journal of comparative neurology.