Cell culture product quality attribute prediction using convolutional neural networks and Raman spectroscopy

Abstract

Advanced process control in the biopharmaceutical industry often lacks real‐time measurements due to resource constraints. Raman spectroscopy and Partial Least Squares (PLS) models are often used to monitor bioprocess cultures in real‐time. In spite of the ease of training, the accuracy of the PLS model is impacted if it is not used to predict quality attributes for the cell lines it is trained on. To address this issue, a deep convolutional neural network (CNN) is proposed for offline modeling of metabolites using Raman spectroscopy. By utilizing asymmetric least squares smoothing to adjust Raman spectra baselines, a generic training data set is created by amalgamating spectra from various cell lines and operating conditions. This data set, combined with their derivatives, forms a two‐dimensional model input. The CNN model is developed and validated for predicting different quality variables against measurements from various continuous and fed‐batch experimental runs. Validation results confirm that the deep CNN model is an accurate generic model of the process to predict real‐time quality attributes, even in experimental runs not included in the training data. This model is robust and versatile, requiring no recalibration when deployed at different sites to monitor various cell lines and experimental runs.

Integration of just‐in‐time learning with variational autoencoder for cell culture process monitoring based on Raman spectroscopy

“…Thus, the need to downsample the Raman features while preserving their statistical attributes becomes imperative. In our previous research (Khodabandehlou et al, 2023), it was demonstrated that downsampling a Raman signal by a factor of 10 preserves its statistical properties. This suggests that downsampling by a factor of less than 10 can also maintain these properties.…”

“…A Raman baseline correction technique was proposed by Eilers and Boelens utilizing asymmetric least square smoothing (Eilers & Boelens, 2005). In this technique, the baseline of the Raman spectra is calculated as follows (Khodabandehlou et al, 2023):…”

“…where y b Δ, , , and γ are the difference operator, the Raman spectra, the baseline, and the penalty factor, respectively. Furthermore, further details on the optimal parameter tuning for the baseline correction method can be found in Khodabandehlou et al (2023).…”

“…While no sample preparation is required with Raman spectroscopy, the sampling frequency of Raman spectra can be high enough (e.g., once per minute) to be reliably used as an alternative to manual sampling through model-based predictions (Khodabandehlou et al, 2023). The predictions are achievable by correlating Raman spectra to process variables using Raman models .…”

“…Specifically, convolutional neural networks (CNNs) can recognize the effective features of Raman spectra through various types of layers (Fukuhara et al, 2019). Given (Khodabandehlou et al, 2023). Despite the promise of deep learning algorithms, it is important to note some limitations.…”