Automatic Zebrafish Egg Phenotype Recognition from Bright-Field Microscopic Images Using Deep Convolutional Neural Network

Abstract

Zebrafish eggs are widely used in biological experiments to study the environmental and genetic influence on embryo development. Due to the high throughput of microscopic imaging, automated analysis of zebrafish egg microscopic images is highly demanded. However, machine learning algorithms for zebrafish egg image analysis suffer from the problems of small imbalanced training dataset and subtle inter-class differences. In this study, we developed an automated zebrafish egg microscopic image analysis algorithm based on deep convolutional neural network (CNN). To tackle the problem of insufficient training data, the strategies of transfer learning and data augmentation were used. We also adopted the global averaged pooling technique to overcome the subtle phenotype differences between the fertilized and unfertilized eggs. Experimental results of a five-fold cross-validation test showed that the proposed method yielded a mean classification accuracy of 95.0% and a maximum accuracy of 98.8%. The network also demonstrated higher classification accuracy and better convergence performance than conventional CNN methods. This study extends the deep learning technique to zebrafish egg phenotype classification and paves the way for automatic bright-field microscopic image analysis.

Deep Learning-Enabled Morphometric Analysis for Toxicity Screening Using Zebrafish Larvae

“…Among different phenotypes, D and U were relatively easy to distinguish, comparable with some early studies. 19,22 Phenotypes like HH, PE, YE, and BS were uniquely different from their normal features, and the variance regularly aggravated with the severity of abnormal phenotypes (i.e., body curvature reduced, yolk size increased). Therefore, the classification results of these phenotypes also had high precision.…”

Translating cuneiform symbols using artificial neural network

“…Now, because of the limitations of available data, image augmentations are exploited to provide a big number of training information. As mentioned in [20][21][22], augmentation strategies of rotations and translations can be utilized. Various training images have been established by applying different rotation and translation processes.…”

Classification of Heart Sounds Using Convolutional Neural Network

“…Therefore, it regularizes the network in order to prevent overfitting. Furthermore, one of the authors of the present study has proven that the strategy of replacing the fully connected layer with a GAP layer is efficient [12]. In this study, we introduced a GAP layer into the CNN model to process medical signals.…”

“…Deep learning [10], especially for convolutional neural network (CNN), has been used to achieve great success in such fields, image detection [11,12], speech recognition [13] and medical image and signal analysis [14]. Recently, many researchers have used deep learning methods to automatically classify heart sounds.…”