Iterative pseudo balancing for stem cell microscopy image classification
Abstract Many critical issues arise when training deep neural networks using limited biological datasets. These include overfitting, exploding/vanishing gradients and other inefficiencies which are exacerbated by class imbalances and can affect the overall accuracy of a model. There is a need to dev...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2024-02-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-024-54993-y |
| _version_ | 1797275148404916224 |
|---|---|
| author | Adam Witmer Bir Bhanu |
| author_facet | Adam Witmer Bir Bhanu |
| author_sort | Adam Witmer |
| collection | DOAJ |
| description | Abstract Many critical issues arise when training deep neural networks using limited biological datasets. These include overfitting, exploding/vanishing gradients and other inefficiencies which are exacerbated by class imbalances and can affect the overall accuracy of a model. There is a need to develop semi-supervised models that can reduce the need for large, balanced, manually annotated datasets so that researchers can easily employ neural networks for experimental analysis. In this work, Iterative Pseudo Balancing (IPB) is introduced to classify stem cell microscopy images while performing on the fly dataset balancing using a student-teacher meta-pseudo-label framework. In addition, multi-scale patches of multi-label images are incorporated into the network training to provide previously inaccessible image features with both local and global information for effective and efficient learning. The combination of these inputs is shown to increase the classification accuracy of the proposed deep neural network by 3 $$\%$$ % over baseline, which is determined to be statistically significant. This work represents a novel use of pseudo-labeling for data limited settings, which are common in biological image datasets, and highlights the importance of the exhaustive use of available image features for improving performance of semi-supervised networks. The proposed methods can be used to reduce the need for expensive manual dataset annotation and in turn accelerate the pace of scientific research involving non-invasive cellular imaging. |
| first_indexed | 2024-03-07T15:09:09Z |
| format | Article |
| id | doaj.art-764fde1458384e088ff3bac9ac93b27f |
| institution | Directory Open Access Journal |
| issn | 2045-2322 |
| language | English |
| last_indexed | 2024-03-07T15:09:09Z |
| publishDate | 2024-02-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj.art-764fde1458384e088ff3bac9ac93b27f2024-03-05T18:43:54ZengNature PortfolioScientific Reports2045-23222024-02-0114111710.1038/s41598-024-54993-yIterative pseudo balancing for stem cell microscopy image classificationAdam Witmer0Bir Bhanu1Department of Bioengineering, University of CaliforniaDepartment of Bioengineering, University of CaliforniaAbstract Many critical issues arise when training deep neural networks using limited biological datasets. These include overfitting, exploding/vanishing gradients and other inefficiencies which are exacerbated by class imbalances and can affect the overall accuracy of a model. There is a need to develop semi-supervised models that can reduce the need for large, balanced, manually annotated datasets so that researchers can easily employ neural networks for experimental analysis. In this work, Iterative Pseudo Balancing (IPB) is introduced to classify stem cell microscopy images while performing on the fly dataset balancing using a student-teacher meta-pseudo-label framework. In addition, multi-scale patches of multi-label images are incorporated into the network training to provide previously inaccessible image features with both local and global information for effective and efficient learning. The combination of these inputs is shown to increase the classification accuracy of the proposed deep neural network by 3 $$\%$$ % over baseline, which is determined to be statistically significant. This work represents a novel use of pseudo-labeling for data limited settings, which are common in biological image datasets, and highlights the importance of the exhaustive use of available image features for improving performance of semi-supervised networks. The proposed methods can be used to reduce the need for expensive manual dataset annotation and in turn accelerate the pace of scientific research involving non-invasive cellular imaging.https://doi.org/10.1038/s41598-024-54993-yDeep learningStem cell microscopyPseudo-labels |
| spellingShingle | Adam Witmer Bir Bhanu Iterative pseudo balancing for stem cell microscopy image classification Scientific Reports Deep learning Stem cell microscopy Pseudo-labels |
| title | Iterative pseudo balancing for stem cell microscopy image classification |
| title_full | Iterative pseudo balancing for stem cell microscopy image classification |
| title_fullStr | Iterative pseudo balancing for stem cell microscopy image classification |
| title_full_unstemmed | Iterative pseudo balancing for stem cell microscopy image classification |
| title_short | Iterative pseudo balancing for stem cell microscopy image classification |
| title_sort | iterative pseudo balancing for stem cell microscopy image classification |
| topic | Deep learning Stem cell microscopy Pseudo-labels |
| url | https://doi.org/10.1038/s41598-024-54993-y |
| work_keys_str_mv | AT adamwitmer iterativepseudobalancingforstemcellmicroscopyimageclassification AT birbhanu iterativepseudobalancingforstemcellmicroscopyimageclassification |