Data Augmentation of Spectral Data for Convolutional Neural Network (CNN) Based Deep Chemometrics
This work addresses chemometrics for pharmaceutical analysis, offering incremental improvements in spectral data processing and model robustness.
The paper tackled predicting drug content in tablets from near infrared spectra using CNNs, achieving results that matched or exceeded optimal PLS models, with data augmentation and EMSC preprocessing yielding the best performance, and CNNs outperforming PLS in extrapolation challenges.
Deep learning methods are used on spectroscopic data to predict drug content in tablets from near infrared (NIR) spectra. Using convolutional neural networks (CNNs), features are ex- tracted from the spectroscopic data. Extended multiplicative scatter correction (EMSC) and a novel spectral data augmentation method are benchmarked as preprocessing steps. The learned models perform better or on par with hypothetical optimal partial least squares (PLS) models for all combinations of preprocessing. Data augmentation with subsequent EMSC in combination gave the best results. The deep learning model CNNs also outperform the PLS models in an extrapolation chal- lenge created using data from a second instrument and from an analyte concentration not covered by the training data. Qualitative investigations of the CNNs kernel activations show their resemblance to wellknown data processing methods such as smoothing, slope/derivative, thresholds and spectral region selection.