Deep Learning: Tackling the challenges of bioimage analysis

  1. Daniël M Pelt  Is a corresponding author
  1. Leiden Institute of Advanced Computer Science, Leiden University, Netherlands


Using multiple human annotators and ensembles of trained networks can improve the performance of deep-learning methods in research.

Main text

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.


  1. Conference
    1. Müller R
    2. Kornblith S
    3. Hinton GE
    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.
    1. Srivastava N
    2. Hinton G
    3. Krizhevsky A
    4. Sutskever I
    5. Salakhutdinov R
    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
    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: December 2, 2020 (version 1)


© 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.


  • 988
    Page views
  • 93
  • 2

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.

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
Deep Learning: Tackling the challenges of bioimage analysis
eLife 9:e64384.

Further reading

    1. Computational and Systems Biology
    2. Genetics and Genomics
    Noushin Hadadi et al.
    Tools and Resources

    Thermal adaptation is an extensively used intervention for enhancing or suppressing thermogenic and mitochondrial activity in adipose tissues. As such, it has been suggested as a potential lifestyle intervention for body weight maintenance. While the metabolic consequences of thermal acclimation are not limited to the adipose tissues, the impact on the rest of the tissues in context of their gene expression profile remains unclear. Here, we provide a systematic characterization of the effects in a comparative multi-tissue RNA sequencing approach following exposure of mice to 10 °C, 22 °C, or 34 °C in a panel of organs consisting of spleen, bone marrow, spinal cord, brain, hypothalamus, ileum, liver, quadriceps, subcutaneous-, visceral- and brown adipose tissues. We highlight that transcriptional responses to temperature alterations exhibit a high degree of tissue-specificity both at the gene level and at GO enrichment gene sets, and show that the tissue-specificity is not directed by the distinct basic gene expression pattern exhibited by the various organs. Our study places the adaptation of individual tissues to different temperatures in a whole-organism framework and provides integrative transcriptional analysis necessary for understanding the temperature-mediated biological programming.

    1. Computational and Systems Biology
    2. Medicine
    Riku Klén et al.
    Research Article

    New SARS-CoV-2 variants, breakthrough infections, waning immunity, and sub-optimal vaccination rates account for surges of hospitalizations and deaths. There is an urgent need for clinically valuable and generalizable triage tools assisting the allocation of hospital resources, particularly in resource-limited countries. We developed and validate CODOP, a machine learning-based tool for predicting the clinical outcome of hospitalized COVID-19 patients. CODOP was trained, tested and validated with six cohorts encompassing 29223 COVID-19 patients from more than 150 hospitals in Spain, the USA and Latin America during 2020-22. CODOP uses 12 clinical parameters commonly measured at hospital admission for reaching high discriminative ability up to nine days before clinical resolution (AUROC: 0·90-0·96), it is well calibrated, and it enables an effective dynamic risk stratification during hospitalization. Furthermore, CODOP maintains its predictive ability independently of the virus variant and the vaccination status. To reckon with the fluctuating pressure levels in hospitals during the pandemic, we offer two online CODOP calculators, suited for undertriage or overtriage scenarios, validated with a cohort of patients from 42 hospitals in three Latin American countries (78-100% sensitivity and 89-97% specificity). The performance of CODOP in heterogeneous and geographically disperse patient cohorts and the easiness of use strongly suggest its clinical utility, particularly in resource-limited countries.