Early and differential diagnosis of Alzheimer's disease (AD) is a problem that puzzled many doctors. Reliable markers in easy-assembling samples are of considerable clinical diagnostic value. In this work, laser Raman spectroscopy (LRS) was developed a new method that potentially allows early and differential diagnosis of AD from the platelet sample. Raman spectra of platelets isolated from different ages of AD transgenic mice and non-transgenic controls were collected and analyzed. Multilayer perceptron networks (MLP) classification method was used to classify spectra and establish the diagnostic models. For differential diagnosis, spectra of platelets isolated from AD, Parkinson's disease (PD) and vascular dementia (VD) mice were also discriminated. Two notable spectral differences at 740 and 1654 cm −1 were revealed in the mean spectrum of platelets isolated from AD transgenic mice and the controls. MLP displayed a powerful ability in the classifying of early, advanced AD and the control group, and in differential diagnosis of PD and advanced AD, as well as VD and advanced AD. The results suggest that platelet detecting by LRS coupled with MLP analysis appears to be an easy and accurate method for early and differential diagnosis of AD. This technique could be rapidly promoted from laboratory to the hospital.

No abstract

No abstract

No abstract

No abstract