A deep-learning model for cancer detection trained on a large number of scanned pathology slides and associated diagnosis labels enables model development without the need for pixel-level annotations.
References
Litjens, G. et al. Sci. Rep. 6, 26286 (2016).
Ehteshami Bejnordi, B. et al. JAMA 318, 2199–210 (2017).
Campanella, G. et al. Nat. Med. 25, 1301–1309 (2019).
Xu, Y., Zhu, J. Y., Chang, E., Lai, M. & Tu, Z. Med. Image Anal. 18, 591–604 (2014).
Tellez, D. et al. IEEE Trans. Med. Imaging 37, 2126–2136 (2018).
Ilse, M., Tomczak, J. M. & Welling, M. Preprint at arXiv https://arxiv.org/abs/1802.04712v4 (2018).
Carbonneau, M. A., Cheplygina, V., Granger, E. & Gagnon, G. Pattern Recog. 77, 329–353 (2018).
Liu, Y. et al. Arch. Pathol. Lab. Med. 143, 859–868 (2019).
Abels, E. et al. J. Pathol. https://doi.org/10.1002/path.5331 (2019).
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Competing interests
Jeroen van der Laak is a member of the scientific advisory boards of Philips (The Netherlands) and ContextVision (Sweden), and receives research funding from Philips (The Netherlands) and from Sectra (Sweden). The remaining authors declare no competing interests.
Rights and permissions
About this article
Cite this article
van der Laak, J., Ciompi, F. & Litjens, G. No pixel-level annotations needed. Nat Biomed Eng 3, 855–856 (2019). https://doi.org/10.1038/s41551-019-0472-6
Published:
Issue Date:
DOI: https://doi.org/10.1038/s41551-019-0472-6
- Springer Nature Limited