There is an urgent clinical need for a fast and effective method for diagnosing Alzheimer’s disease (AD). The identification of AD in its most initial stages, at which point treatment could provide maximum therapeutic benefits, is not only likely to slow down disease progression but to also potentially provide a cure. However, current clinical detection is complicated and requires a combination of several methods based on significant clinical manifestations due to widespread neurodegeneration. As such, Raman spectroscopy with machine learning is investigated as a novel alternative method for detecting AD in its earliest stages. Here, blood serum obtained from rats fed either a standard diet or a high-fat diet was analyzed. The high-fat diet has been shown to initiate a pre-AD state. Partial least squares discriminant analysis combined with receiver operating characteristic curve analysis was able to separate the two rat groups with 100% accuracy at the donor level during external validation. Although further work is necessary, this research suggests there is a potential for Raman spectroscopy to be used in the future as a successful method for identifying AD early on in its progression, which is essential for effective treatment of the disease.
Towards development of a novel screening method for identifying Alzheimer’s disease risk: Raman spectroscopy of blood serum and machine learning.
