en-GBde-DEes-ESfr-FR

Designing and evaluating medical deep learning systems


A new Perspective article in Nature Reviews Cancer discusses reasons for the moderate progress and describes remedies designed to facilitate the transition to the clinic.

Can better design of deep learning studies lead to the faster transformation of medical practices? According to the authors of Designing deep learning studies in cancer diagnostics in Nature Reviews Cancer’s latest issue, the answer is yes.

- We propose several protocol items that should be defined before evaluating the external cohort says the paper’s first writer Andreas Kleppe at Institute for Cancer Diagnostics and Informatics at Oslo University Hospital.

- In this way, the evaluation becomes rigorous and more reliable. Such evaluations would make it much clearer which systems are likely to work well in clinical practice, and these systems should be further assessed in phase III randomised clinical trials.

Slow implementation is partly a natural consequence of the time needed to evaluate and adapt systems affecting patient treatment. However, many studies assessing well-functioning systems are at high risk of bias.

According to Kleppe, even among the seemingly best studies that evaluate external cohorts, few predefine the primary analysis. Adaptations of the deep learning system, patient selection or analysis methodology can make the results presented over-optimistic.

The frequent lack of stringent evaluation of external data is of particular concern. Some systems are developed or evaluated on too narrow or inappropriate data for the intended medical setting. The lack of a well-established sequence of evaluation steps for converting promising prototypes into properly evaluated medical systems limits deep learning systems' medical utilisation.

Millions of adjustable parameters

Deep learning facilitates utilisation of large data sets through direct learning of correlations between raw input data and target output, providing systems that may use intricate structures in high-dimensional input data to model the association with the target output accurately. Whereas supervised machine learning techniques traditionally utilised carefully selected representations of the input data to predict the target output, modern deep learning techniques use highly flexible artificial neural networks to correlate input data directly to the target outputs.

The relations learnt by such direct correlation will often be true but may sometimes be spurious phenomena exclusive to the data utilised for learning. The millions of adjustable parameters make deep neural networks capable of performing correctly in training sets even when the target outputs are randomly generated and, therefore, utterly meaningless.

Design and evaluation challenges

The high capacity of neural networks induces severe challenges for designing and developing deep learning systems and validating their performance in the intended medical setting. An adequate clinical performance will only be possible if the system has good generalisability to subjects not included in the training data.

The design challenges involve selecting appropriate training data, such as representativeness of the target population. It also includes modelling questions such as how the variation of training data may be artificially increased without jeopardising the relationship between input data and target outputs in the training data.

The validation challenge includes verifying that the system generalises well. For example, does it perform satisfactorily when evaluated on relevant patient populations at new locations and when input data are obtained using differing laboratory procedures or alternative equipment? Moreover, deep learning systems are typically developed iteratively, with repeated testing and various selection processes that may bias results. Similar selection issues have been recognised as a general concern for the medical literature for many years.

Thus, when selecting design and validation processes for diagnostic deep learning systems, one should focus on the generalisation challenges and prevent more classical pitfalls in data analysis.

- To achieve good performance for new patients, it is crucial to use various training data. Natural variation is always essential, but so is introducing artificial variation. These types of variation complement each other and facilitate good generalisability, says Kleppe.

Nature Reviews Cancer
Designing deep learning studies in cancer diagnostics
Andreas Kleppe, Ole-Johan Skrede, Sepp De Raedt, Knut Liestøl, David J. Kerr & Håvard E. Danielsen
Nature Reviews Cancer (2021)Cite this article
Attached files
  • bilder-32.jpeg
  • 9110anonymifisert202.jpg
Regions: Europe, Norway
Keywords: Health, Medical, Science, Mathematics

Testimonials

We have used AlphaGalileo since its foundation but frankly we need it more than ever now to ensure our research news is heard across Europe, Asia and North America. As one of the UK’s leading research universities we want to continue to work with other outstanding researchers in Europe. AlphaGalileo helps us to continue to bring our research story to them and the rest of the world.
Peter Dunn, Director of Press and Media Relations at the University of Warwick
AlphaGalileo has helped us more than double our reach at SciDev.Net. The service has enabled our journalists around the world to reach the mainstream media with articles about the impact of science on people in low- and middle-income countries, leading to big increases in the number of SciDev.Net articles that have been republished.
Ben Deighton, SciDevNet
AlphaGalileo is a great source of global research news. I use it regularly.
Robert Lee Hotz, LA Times

We Work Closely With...


  • BBC
  • The Times
  • National Geographic
  • The University of Edinburgh
  • University of Cambridge
Copyright 2021 by DNN Corp Terms Of Use Privacy Statement