Back

Leveraging Transfer Learning For High-Accuracy Phenotypic Screening In Zebrafish Image Analysis

Jayaraman, V. U.; Medishetti, R.; Ghosh, S.; Chatti, K.; Uppada, M. K.; Oruganti, S.

2025-01-24 bioinformatics
10.1101/2025.01.21.634041 bioRxiv
Show abstract

This paper presents a method for classifying zebrafish images captured before and after drug administration. Leveraging the power of transfer learning and fine-tuning, the approach effectively overcomes the challenges of limited datasets in biomedical imaging. By employing a pre-trained convolutional neural network (CNN) as the base model, transfer learning allows us to utilize learned features from large-scale image datasets, significantly reducing training time and computational resources. Fine-tuning specific layers of the model on our zebrafish dataset further enhances its ability to detect subtle visual differences induced by drug administration. The proposed approach achieves high accuracy in classifying zebrafish images, demonstrating its potential as a reliable tool for analysing phenotypic changes due to pharmacological interventions. This model could be instrumental in accelerating drug discovery and research in zebrafish-based assays, offering a scalable and efficient solution for image-based biomedical analysis.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.