A Novel Cell Density Prediction Design using Optimal Deep Learning with Salp Swarm Algorithm

Abstract

Cell density prediction can be defined as the process of predicting the number of cells in a given quantity of a culture or cell suspension. It is considered a common practice in cell biology since cell density had a significant impact on cell behavior and can be utilized for monitoring the health and growth of cell culture. Precise prediction of cell density was significant for a range of applications in cell biology., which includes bioprocessing, cell-based assays, and cell culture. Therefore, this article develops a novel Cell Density Prediction design using Optimal Deep Learning with Salp Swarm Algorithm (CDP-ODLSSA) technique. The presented CDP-ODLSSA technique predicts the cell densities accurately on the images of cell suspensions or cultures. To do so, the presented CDP-ODLSSA technique employs Long Short Term Memory-Autoencoder (LSTM-AE) model for prediction of cell densities. In addition, the hyperparameter tuning of the LSTM-AE model takes place by the use of Salp Swarm Algorithm (SSA). For experimental validation of the CDP-ODLSSA technique, a wide range of simulations was taken place. The obtained values highlighted the superiority of the CDP-ODLSSA technique compared to other approaches.