Online Short Wave Infrared Hyperspectral AI-Imaging System (SWIRhyperspectral AI-Imaging) for measuring cell culture material, process and quality attributes in mammalian NISTmAb-producing NISTCHO cell cultures

Within the biopharmaceutical sector there exist a need for a contactless sensor which can accurately predict quality attributes in real time for precise feedback control of a bioreactor environment. The development of a short wave infrared hyperspectral imaging (SWIR-HSI) system allows for an efficient, fully contactless collection of large spectral datasets for metabolite quantification. Here, we build on the Phase I development of an interpretable AI/deep learning system using a convolution metabolite regression (CMR) approach for glucose, lactate into a convolution attribute regression (CAR) approach for glycosylation, aggregation and charge variants using label-free contactless hyperspectral images in the SWIR spectral range of NISTCHO cells producing NISTmAb grown in a 3L bioreactor. We propose generating a dataset of >10,000 spectral images obtained through a proprietary optically transparent microfluidic chip with flowing CHO cells and spent media online in order to predict NISTCHO cell viability, antibody titers, galacosylation, a-fucosylation, soluble aggregates, and charge variants using the CAR model and a training set comprising ground truth readings of each of these material attributes and process parameters in real-time. Collectively, this work will advance the safe and efficient adoption of contactless AI/deep learning sensing systems for fine control of a variety of bioreactor environments.