12
 OCT

12. October 2022

How Deep Learning is Used for Cell Counting

The process of cell counting is important to biomedical research as it can reveal valuable information about single cells, populations of cells, and cell dynamics and furthermore it is the basis of more sophisticated tissue cytometry workflows. Cell counting relies on robust biomedical imaging techniques to accurately and efficiently be performed. Cell counting (or nuclei segmentation) based on deep learning algorithms is a type of artificial intelligence that enhances the accuracy of nuclei segmentation.

This article serves as an introduction to algorithmic cell counting via deep learning. We will briefly outline what a deep neural network is and how it applies to applications of cell counting deep learning.

What is a deep neural network for cell counting deep learning?

A deep neural network (DNN) is a more complex version of artificial neural networks used by advanced computer technologies to learn and model information in more understandable ways [1]. A DNN is described as having multiple layers that executes model training to improve accuracy while it is performing the task the computer is programmed to accomplish for researchers.

A DNN is very useful for cell counting. TissueGnostics have developed a DNN to perform nuclear segmentation in dense tissue microenvironments, which was previously a hurdle to overcome due to a high density of cells making identification of single cells challenging [2]. DNN amplify cell counting abilities by continuously improving models in order to output the most accurate data it can possibly create at that moment in time.

TissueGnostics as a leading developer of cell counting deep learning

TissueGnostics is a leading tissue cytometry solutions provider that delivers products and services designed to automate procedures including scanning of IF or IHC processed samples (tissue sections, biopsies, TMAs, cultured cells…) as well as high end image analysis (cell counting/segmentation via deep learning´, tissue classification, spatial analysis.

More information about the wide selection of platforms and analysis solutions available can be found on our website. If you are interested in gaining more in-depth knowledge about cell counting deep learning and our services, please reach out to a member of the TissueGnostics team as soon as possible, and we would be happy to answer your questions.

Reference

1. BMC Blogs. (2020). What’s a Deep Neural Network? Deep Nets Explained. [online] Available at: https://www.bmc.com/blogs/deep-neural-network/
2. Felicitas (2022). Deep Neural Network | DNN | Nuclei Detection. [online] Tissuegnostics.com. Available at: https://tissuegnostics.com/products/contextual-image-analysis/ai-solutions-in-strataquest/if-dnn
3. Morelli, R., Clissa, L., Amici, R., Cerri, M., Hitrec, T., Luppi, M., Rinaldi, L., Squarcio, F. and Zoccoli, A. (2021). Automating cell counting in fluorescent microscopy through deep learning with c-ResUnet. Scientific Reports, [online] 11(1). Available at: https://www.nature.com/articles/s41598-021-01929-5
4. Dmitry Urukov (2020). Real-time cell counting in microscopy images using Neural Networks. [online] Medium. Available at: https://medium.com/analytics-vidhya/real-time-cell-counting-in-microscopy-images-with-neural-networks-d630c2a5b6c4
5. Wang, D., Hwang, M., Jiang, W.-C., Ding, K., Chang, H.C. and Hwang, K.-S. (2021). A deep learning method for counting white blood cells in bone marrow images. BMC Bioinformatics, [online] 22(S5). Available at: https://bmcbioinformatics.biomedcentral.com/articles/10.1186/s12859-021-04003-z

Are you also interested in How Cell Counting Algorithms Work?

or in How to Carry Out Complex Characterization of the Tissue Microenvironment?