Complementary performances of convolutional and capsule neural networks on classifying microfluidic images of dividing yeast cells

• Published in: PloS one Vol. 16; no. 3; p. e0246988
• Main Authors: Ghafari, Mehran, Clark, Justin, Guo, Hao-Bo, Yu, Ruofan, Sun, Yu, Dang, Weiwei, Qin, Hong
• Format: Journal Article
• Language: English
• Published: United States Public Library of Science 17.03.2021; Public Library of Science (PLoS)
• Subjects: Accuracy; Aging; Architecture; Biology and Life Sciences; Classification; Computer and Information Sciences; Computer programs; Computer science; Datasets; Deep learning; Drafting software; Editing; Electronic mail; Engineering; Engineering and Technology; Environmental science; Genetic aspects; Geology; Handwriting; Identification and classification; Image classification; Image resolution; Machine learning; Medicine; Methods; Microfluidics; Neural networks; Physiological aspects; Research and Analysis Methods; Reviews; Software; Training; Yeast; Yeast fungi; Yeasts; United States; United States > US; Texas; Tennessee
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0246988

Microfluidic-based assays have become effective high-throughput approaches to examining replicative aging of budding yeast cells. Deep learning may offer an efficient way to analyze a large number of images collected from microfluidic experiments. Here, we compare three deep learning architectures to classify microfluidic time-lapse images of dividing yeast cells into categories that represent different stages in the yeast replicative aging process. We found that convolutional neural networks outperformed capsule networks in terms of accuracy, precision, and recall. The capsule networks had the most robust performance in detecting one specific category of cell images. An ensemble of three best-fitted single-architecture models achieves the highest overall accuracy, precision, and recall due to complementary performances. In addition, extending classification classes and data augmentation of the training dataset can improve the predictions of the biological categories in our study. This work lays a useful framework for sophisticated deep-learning processing of microfluidic-based assays of yeast replicative aging.