Non-negative spectral unmixing of Raman spectra for determination of dopamine concentrations in rat brain tissue

In fifteen words or less, explain the significance of this contribution (Novel Aspect).: Developing novel techniques for neurological disease diagnosis using Raman spectroscopy and Machine learning.

Abstract Text:

Dopamine is an essential neurochemical that modulates several physiological and behavioral processes in the brain. It is produced by dopaminergic neurons found in the midbrain region, primarily in the substantia nigra and the ventral tegmental area (VTA). Neurons from the VTA project to the prefrontal cortex, and play an important role in motivation, reward processing, and cognition. Accurate determination of the distributions of dopamine in these different regions of the brain provides insight into neurochemical dynamics, which could help improve diagnosis of neurological diseases, such as depression and Parkinson’s disease. We will present the integration of Raman spectroscopy (a non-destructive technique) with an advanced spectral unmixing method, Non-Negative Matrix Factorization (NMF), and similarity analysis to determine the presence and relative abundance of dopamine in tissue slices from rat brains that include the substantia nigra and cortex regions. The Raman spectra of the two regions of interest in the brain tissue are complex, with overlapping contributions from the tissue, neurochemicals, and other biological components present. This spectral complexity necessitates the use of NMF for determination of the presence of dopamine. Cosine similarity analysis is performed between the unmixed components and a reference dopamine spectrum to qualitatively and quantitatively assess dopamine concentrations in the substantia nigra and cortex regions of the brain.