Deep Learning: Tackling the challenges of bioimage analysis

Using multiple human annotators and ensembles of trained networks can improve the performance of deep-learning methods in research.
  1. Daniël M Pelt  Is a corresponding author
  1. Leiden Institute of Advanced Computer Science, Leiden University, Netherlands

Deep learning has shown promising results in a wide range of imaging problems in recent years (LeCun et al., 2015), and has the potential to help researchers by automating the analysis of various kinds of biological images (Ronneberger et al., 2015). However, many deep-learning methods require a large amount of 'training data' in order to produce useful results, and this is often not available for bioimage analysis. There is, therefore, a need for deep-learning methods that can make the most from a limited amount of training data. Now, in eLife, Robert Blum (University Hospital Würzburg), Christoph Flath (University of Würzburg) and colleagues – including Dennis Segebarth and Matthias Griebel as joint first authors – provide guidance on how to do this in bioimage analysis (Segebarth et al., 2020).

A common approach to applying deep learning to image analysis involves 'convolutional neural networks': these networks take an input image (such as a microscopy image) and perform many mathematical operations on it to produce an output image (such as a corresponding image with interesting features annotated). A convolutional neural network is characterized by a set of 'learnable parameters', which have to be set to the correct values for the network to perform a given task. The act of finding the correct values for these parameters is called 'training', and several different training techniques are used in practice.

In supervised learning, training is performed using a set of input images and target output images, and the learnable parameters are iteratively adjusted until the output images produced by the network match the target images. It is important to note that supervised learning involves a large amount of randomness, and that training multiple networks using the same data will result in different networks that produce (slightly) different output images.

In bioimage analysis, a common task is to annotate certain structures in images produced by techniques such as microscopy, cryo-EM or X-ray tomography (Meijering et al., 2016). However, the complicated nature of biological images means that this annotation often has to be done by a human expert, which is time-consuming, labour-intensive and subjective (Figure 1A). Supervised deep learning could provide a way to automate the annotation process, reducing the burden on human experts and enabling analysis of a significantly larger set of images. However, annotating the input images needed to train the network also requires a significant amount of time and effort from a human expert.

Different ways to train a convolutional neural network.

Segebarth et al. compare three techniques for training convolutional neural networks to analyze bioimages. (A) In the standard approach a single human expert annotates images for training a single network. (B) In a second approach multiple human experts annotate the same images, and consensus images are used for training: this improves the objectivity of the trained network. (C) In a third approach, a technique called model ensembling is added to the second approach, meaning that multiple networks are trained with the same consensus images: this improves the reliability of the results.

Several properties of a trained network are important for real-world applications. The first is its objectivity, referring to a lack of influences from the subjective nature of human annotations. The second is its reliability, meaning that the trained network should consistently annotate similar features in the same way. The third is its validity, referring to the truthfulness of the network output (that is, did we annotate what we intended to?). In practice, it is often the case that objectivity, reliability and validity are difficult to achieve when the amount of training data is limited.

Various approaches to improve the objectivity, reliability and validity of convolutional neural networks have been proposed. Some involve adapting the structure of the network themselves by, for example, reducing the number of learnable parameters (Pelt and Sethian, 2018), and some involve adapting the training method by, for example, randomly ignoring parts of the network during training (Srivastava et al., 2014). A different approach is to focus on the training data used in supervised learning. Given a network structure and a training method, how can the training data set be optimized to improve objectivity, reliability and validity? In other words, given the time-consuming, labour-intensive and subjective nature of manual annotation, how can a limited period of time from human experts be best utilized to produce a training data set? These questions are currently the subject of active research.

Segebarth et al. investigate two techniques for improving the objectivity, reliability and validity of trained convolutional neural networks in bioimage analysis. First, they investigate the use of multiple human experts to annotate the same set of training images (Figure 1B). The different annotations of each input image are then combined to create a consensus target output image. Since each human expert has their own intended and unintended biases, networks that are trained with data from a single human expert might include the biases of the expert. Using consensus images from multiple experts during training can improve the objectivity of the resulting networks by removing these biases from the training data.

The second technique is to train multiple convolutional neural networks using the same training data set, and then combine the results when the networks are used to analyse new images (Figure 1C). This technique, called model ensembling, has already proven successful in a wide range of applications (Krizhevsky et al., 2017). Model ensembling is based on the randomness involved in training described above: because of this randomness, each trained network will be implicitly biased in their results. By combining the output of multiple networks, these biases are effectively removed, resulting in more reliable results.

A key contribution of Segebarth et al. was to perform extensive experiments on real images and show that the use of consensus images and model ensembles does indeed improve objectivity, reliability and validity. This provides a recipe for optimizing the generation of training data and for making efficient use of the available data, although this recipe still requires a significant amount of human expert time since each image has to be annotated by multiple experts. The results could also help researchers trying to understand how biases affect trained networks, which could lead to improved network structures and training approaches (Müller et al., 2019). And although many questions and challenges remain, the work of Segebarth et al. represents an important step forward in the effort to make the use of deep learning in bioimage analysis feasible.

References

  1. Conference
    1. Müller R
    2. Kornblith S
    3. Hinton GE
    (2019)
    When does label smoothing help?
    Advances in Neural Information Processing Systems. pp. 4694–4703.
  2. Book
    1. Ronneberger O
    2. Fischer P
    3. Brox T
    (2015) U-net: Convolutional networks for biomedical image segmentation
    In: Navab N, Hornegger J, wells W, Frango A, editors. International Conference on Medical Image Computing and Computer-Assisted Intervention. Springer. pp. 234–241.
    https://doi.org/10.1007/978-3-319-24574-4_28
    1. Srivastava N
    2. Hinton G
    3. Krizhevsky A
    4. Sutskever I
    5. Salakhutdinov R
    (2014)
    Dropout: a simple way to prevent neural networks from overfitting
    Journal of Machine Learning Research 15:1929–1958.

Article and author information

Author details

  1. Daniël M Pelt

    Daniël M Pelt is in the Leiden Institute of Advanced Computer Science, Leiden University, Leiden, Netherlands

    For correspondence
    Daniel.Pelt@cwi.nl
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-8253-0851

Publication history

  1. Version of Record published:

Copyright

© 2020, Pelt

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 1,155
    views
  • 111
    downloads
  • 9
    citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Daniël M Pelt
(2020)
Deep Learning: Tackling the challenges of bioimage analysis
eLife 9:e64384.
https://doi.org/10.7554/eLife.64384

Further reading

    1. Cell Biology
    2. Computational and Systems Biology
    Sarah De Beuckeleer, Tim Van De Looverbosch ... Winnok H De Vos
    Research Article

    Induced pluripotent stem cell (iPSC) technology is revolutionizing cell biology. However, the variability between individual iPSC lines and the lack of efficient technology to comprehensively characterize iPSC-derived cell types hinder its adoption in routine preclinical screening settings. To facilitate the validation of iPSC-derived cell culture composition, we have implemented an imaging assay based on cell painting and convolutional neural networks to recognize cell types in dense and mixed cultures with high fidelity. We have benchmarked our approach using pure and mixed cultures of neuroblastoma and astrocytoma cell lines and attained a classification accuracy above 96%. Through iterative data erosion, we found that inputs containing the nuclear region of interest and its close environment, allow achieving equally high classification accuracy as inputs containing the whole cell for semi-confluent cultures and preserved prediction accuracy even in very dense cultures. We then applied this regionally restricted cell profiling approach to evaluate the differentiation status of iPSC-derived neural cultures, by determining the ratio of postmitotic neurons and neural progenitors. We found that the cell-based prediction significantly outperformed an approach in which the population-level time in culture was used as a classification criterion (96% vs 86%, respectively). In mixed iPSC-derived neuronal cultures, microglia could be unequivocally discriminated from neurons, regardless of their reactivity state, and a tiered strategy allowed for further distinguishing activated from non-activated cell states, albeit with lower accuracy. Thus, morphological single-cell profiling provides a means to quantify cell composition in complex mixed neural cultures and holds promise for use in the quality control of iPSC-derived cell culture models.

    1. Computational and Systems Biology
    Franck Simon, Maria Colomba Comes ... Herve Isambert
    Tools and Resources

    Live-cell microscopy routinely provides massive amounts of time-lapse images of complex cellular systems under various physiological or therapeutic conditions. However, this wealth of data remains difficult to interpret in terms of causal effects. Here, we describe CausalXtract, a flexible computational pipeline that discovers causal and possibly time-lagged effects from morphodynamic features and cell–cell interactions in live-cell imaging data. CausalXtract methodology combines network-based and information-based frameworks, which is shown to discover causal effects overlooked by classical Granger and Schreiber causality approaches. We showcase the use of CausalXtract to uncover novel causal effects in a tumor-on-chip cellular ecosystem under therapeutically relevant conditions. In particular, we find that cancer-associated fibroblasts directly inhibit cancer cell apoptosis, independently from anticancer treatment. CausalXtract uncovers also multiple antagonistic effects at different time delays. Hence, CausalXtract provides a unique computational tool to interpret live-cell imaging data for a range of fundamental and translational research applications.