Depth IR spectroscopic data resolution improvement for antibiotics component analysis in critically ill elderly patients

Abstract Critically ill patients in intensive care units are vulnerable to the bacterial infections, especially if those patients are elderly. The effective utilization of the appropriate drug can reduce the mortality rates in these patients. Overlap bands often appear in applications of infrared (IR) spectroscopy, for instance in the identification of the unknown material or drugs. This paper considers the problem of noisy and overlapped IR spectroscopic data resolution improvement. A resolution improvement approach with ridgelet transform regularization for IR spectroscopic data and total variation regularization for the apparatus response function. Moreover, the split Bregman method is exploited to solve the resulting minimization problem. It is computationally simple and suitable for implementation on small computers with less memory requirements. Simulation experimental results demonstrates the excellent performance of the proposed approach at noise suppression and spectral detail preserving. The proposed method can remove the random noise and improve the spectroscopic data resolution, thus leading the high-resolution IR spectroscopic data a more efficient tool for component analysis of the unknown antibiotics or drug in critically ill elderly patients.