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

Protein production in the biopharmaceutical industry necessitates the utilization of multiple analytical techniques and control methodologies to ensure both safety and consistency. To facilitate real-time monitoring and control of cell culture processes, Raman spectroscopy has emerged as a versatile analytical technology. This technique, categorized as a Process Analytical Technology, employs chemometric models to establish correlations between Raman signals and key variables of interest. One notable approach for achieving real-time monitoring is through the application of just-in-time learning (JITL), an industrial soft sensor modeling technique that utilizes Raman signals to estimate process variables promptly. The conventional Raman-based JITL method relies on the K-nearest neighbor (KNN) algorithm with Euclidean distance as the similarity measure. However, it falls short of addressing the impact of data uncertainties. To rectify this limitation, this study endeavors to integrate JITL with a variational autoencoder (VAE). This integration aims to extract dominant Raman features in a nonlinear fashion, which are expressed as multivariate Gaussian distributions. Three experimental runs using different cell lines were chosen to compare the performance of the proposed algorithm with commonly utilized methods in the literature. The findings indicate that the VAE-JITL approach consistently outperforms partial least squares, convolutional neural network, and JITL with KNN similarity measure in accurately predicting key process variables.

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

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.

Machine learning-based model predictive controller design for cell culture processes.

The biopharmaceutical industry continuously seeks to optimize the critical quality attributes to maintain the reliability and cost-effectiveness of its products. Such optimization demands a scalable and optimal control strategy to meet the process constraints and objectives. This work uses a model predictive controller (MPC) to compute an optimal feeding strategy leading to maximized cell growth and metabolite production in fed-batch cell culture processes. The lack of high-fidelity physics-based models and the high complexity of cell culture processes motivated us to use machine learning algorithms in the forecast model to aid our development. We took advantage of linear regression, the Gaussian process and neural network models in the MPC design to maximize the daily protein production for each batch. The control scheme of the cell culture process solves an optimization problem while maintaining all metabolites and cell culture process variables within the specification. The linear and nonlinear models are developed based on real cell culture process data, and the performance of the designed controllers is evaluated by running several real-time experiments.

Subject-specific estimation of central aortic blood pressure via system identification: preliminary in-human experimental study.

This paper demonstrates preliminary in-human validity of a novel subject-specific approach to estimation of central aortic blood pressure (CABP) from peripheral circulatory waveforms. In this "Individualized Transfer Function" (ITF) approach, CABP is estimated in two steps. First, the circulatory dynamics of the cardiovascular system are determined via model-based system identification, in which an arterial tree model is characterized based on the circulatory waveform signals measured at the body's extremity locations. Second, CABP waveform is estimated by de-convolving peripheral circulatory waveforms from the arterial tree model. The validity of the ITF approach was demonstrated using experimental data collected from 13 cardiac surgery patients. Compared with the invasive peripheral blood pressure (BP) measurements, the ITF approach yielded significant reduction in errors associated with the estimation of CABP, including 1.9-2.6 mmHg (34-42 %) reduction in BP waveform errors (p < 0.05) as well as 5.8-9.1 mmHg (67-76 %) and 6.0-9.7 mmHg (78-85 %) reductions in systolic and pulse pressure (SP and PP) errors (p < 0.05). It also showed modest but significant improvement over the generalized transfer function approach, including 0.1 mmHg (2.6 %) reduction in BP waveform errors as well as 0.7 (20 %) and 5.0 mmHg (75 %) reductions in SP and PP errors (p < 0.05).

Comparative study on tube-load modeling of arterial hemodynamics in humans.

In this paper, we assess the validity of two alternative tube-load models for describing the relationship between central aortic and peripheral arterial blood pressure (BP) waveforms in humans. In particular, a single-tube (1-TL) model and a serially connected two-tube (2-TL) model, both terminated with a Windkessel load, are considered as candidate representations of central aortic-peripheral arterial path. Using the central aortic, radial and femoral BP waveform data collected from eight human subjects undergoing coronary artery bypass graft with cardiopulmonary bypass procedure, the fidelity of the tube-load models was quantified and compared with each other. Both models could fit the central aortic-radial and central aortic-femoral BP waveform pairs effectively. Specifically, the models could estimate pulse travel time (PTT) accurately, and the model-derived frequency response was also close to the empirical transfer function estimate obtained directly from the central aortic and peripheral BP waveform data. However, 2-TL model was consistently superior to 1-TL model with statistical significance as far as the accuracy of the central aortic BP waveform was concerned. Indeed, the average waveform RMSE was 2.52 mmHg versus 3.24 mmHg for 2-TL and 1-TL models, respectively (p < 0.05); the r² value between measured and estimated central aortic BP waveforms was 0.96 and 0.93 for 2-TL and 1-TL models, respectively (p < 0.05). We concluded that the tube-load models considered in this paper are valid representations that can accurately reproduce central aortic-radial/femoral BP waveform relationships in humans, although the 2-TL model is preferred if an accurate central aortic BP waveform is highly desired.