
**Detection of Alzheimer's Disease by Machine Learning-assisted Vibrational
Spectroscopy in Cerebrospinal Fluid**
Dr. Laura Arevalo
Nanoengineering, CIC nanoGUNE
Nowadays, the diagnosis of Alzheimer’s disease (AD) is a complex process
that involves several clinical tests such as neurological evaluation or brain
scans [1]. Cerebrospinal Fluid (CSF), as it is in direct contact with the
brain, is used to find biomarkers of amyloid β pathology and tau pathology.
In this talk, a new method of detection of biomarkers in CSF related with the
Alzheimer’s disease is described; the methodology is based on vibrational
spectroscopy analysis through machine learning prediction model [2].
Vibrational spectroscopy provides the entire biochemical composition of the
CSF, in that way the detection of small changes typical of the AD can be
ascertained. Infrared absorption and Raman spectra of CSF samples acquired
from 22 volunteers were measured. A dataset with 610 spectra were analysed
within a Machine Learning Framework. We found that a logistic regression model
can discriminate between healthy and sick patients with a precision of 98%
when the input for the model is the combination of the two types of spectral
data. Our methodology shows high discriminative capabilities and is a proof of
concept of an alternative and accurate tool for the diagnosis of AD.
**References **
[1] Dubois, B., Villain, N., Frisoni, G. B., Rabinovici, G. D., Sabbagh, M.,
Cappa, S., ... & Feldman, H. H. (2021). Clinical diagnosis of Alzheimer's
disease: recommendations of the International Working Group. The Lancet
Neurology, 20(6), 484-496.
[2] Baker, M. J., Byrne, H. J., Chalmers, J., Gardner, P., Goodacre, R.,
Henderson, A., ... & Sule-Suso, J. (2018). Clinical applications of infrared
and Raman spectroscopy: state of play and future challenges. Analyst, 143(8),
1735-1757.