An overview of deep learning-based cell image analysis. A typical analysis pipeline consists of a retraining module and an inference module: the inference module produces the estimated metrics directly. However, as the experimental setup changes, the parameters of the deep learning model must be retrained. attributed to him: intelligent computing (2022). DOI: 10.34133/2022/9861263
The cell is the basic structural and functional unit of life, with different sizes, shapes and densities. There are many different physiological and pathological factors that influence these parameters. It is therefore very important for biomedical and pharmaceutical research to study the properties of cells.
Traditionally, researchers have observed samples of cells directly through microscopes to study the morphological changes of the cells. In recent years, with the development of computer science and Artificial intelligenceDeep learning can now be combined with cell analysis methods. This can replace researchers’ direct observation under a microscope and manual interpretation of images, which improves research efficiency and accuracy.
An increasing number of algorithms based on deep learning have been developed to enable cell image analysis, mainly to address three main tasks:
- hash. To identify meaningful elements or features, the image is divided into several parts using deep learning. Cell segmentation is the basic premise for the identification, computation, tracking, and morphological analysis of cell images;
- tracking. That is, after segmentation of cell images, cell behavior is monitored for the entire spectrum. Living cells contain a lot of information about Organismthe dynamic characteristics of cells, especially morphological changes, can reflect the health status of the organism in the disease state and physiological processesSuch as immune responsewound healing, proliferation of cancer cells and metastases, etc.
- classification. Classification of cell morphological features based on the extracted parameters often serves as a post-hoc analysis task to examine cell phenotype and profiling.
For the three critical tasks mentioned above, I published a review article in the journal intelligent computing It discusses in depth the advances in deep learning techniques.
Unlike traditional computer vision techniques, A deep neural network (DNN) can automatically produce representations that are more efficient than hand-made ones by learning from a large-scale dataset. In cell images, deep learning-based methods also show promising results in cell segmentation and tracking,” the authors said. Such successful applications demonstrate the ability of DNNs to extract high-level features and highlight the potential ability of using deep learning to reveal the more complex laws of life behind cellular patterns.”
In addition, the authors also discuss challenges and opportunities related to deep learning methods in cellular image processing. The authors said, “Deep learning has demonstrated an amazing ability to perform analysis of cell images. However, there is still a significant performance gap between deep learning algorithms in academic research and practical applications.” Currently, there are challenges and opportunities in three aspects: the amount of data, Data qualitydata confidence:
- Deep learning using a small but expensive data set. Generating a large sized cell image dataset is a tedious task. This is because cell images require knowledgeable biologists to map labels picture by picture. The size of cell image datasets is often limited by the difficulty of annotation.
- Deep learning with noisy and unbalanced labels. The annotation quality of cell image datasets is highly dependent on the professional skills of humans, which leads to label confusion and label imbalance. Label noise is introduced by assigning incorrect or incomplete labels to training images. Label imbalance occurs due to annotation preference, in which the numbers of images labeled for different categories are completely unbalanced.
- Perceived uncertainty of analyzing cellular images. Uncertainty-aware learning is critical for applications of deep learning in biological scenarios. It is impossible for a simple neural network to detect new phenotypes without a mechanism to confidently reflect classification results.
Using deep learning, scientists are exploring new techniques to improve cell image analysis. More effective solutions will be proposed in the future, and Deep learning And the Biomedical research will be more integrated.
Junde Xu et al, Deep Learning in Cellular Image Analysis, intelligent computing (2022). DOI: 10.34133/2022/9861263
Introduction of intelligent computing
the quote: How Deep Learning Enables Cell Image Analysis (2022, November 21) Retrieved November 21, 2022 from https://phys.org/news/2022-11-deep-empowers-cell-image-analysis.html
This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no part may be reproduced without written permission. The content is provided for informational purposes only.