Performance evaluation of temporal features for detection of mild cognitive impairment: An fNIRS study

Abstract

A decline in cognition characterized by the deficit in memory, judgment, and language is known as mild cognitive impairment (MCI). This deficiency has great likelihood of developing Alzheimer’s disease (AD). Hence, the recognition of MCI or an early stage of AD is critical at earliest for evaluation and treatment. In this study, functional near-infrared spectroscopy (fNIRS) was used to measure the concentration changes of oxygenated hemoglobin (ΔHbO) from the pre-frontal cortex of six healthy controls (HC) and six MCI patients during a working memory task. Various temporal features of ΔHbO, such as mean, slope, kurtosis, skewness, and peak-values, were tested for distinguishing MCI patients from HCs. We further investigated the impact of intervention (i.e., acupuncture treatment) on MCI patients (i.e., MCI-1: before intervention, MCI-2: after intervention). Both groups went through fNIRS recording procedure with a gap of 45 days. Before the second fNIRS measurement, only the patient group underwent interventions twice per week for six weeks. The results showed that the mean ΔHbO response of MCI-1 was significantly less than (p < 0.05) that of HC but similar (p > 0.05) with MCI-2. Out of tested temporal features, signal mean and peak values of HC were significantly different (p = 0.0409, p = 0.0430) in comparison to MCI-1 but no significant difference (p = 0.4885, p = 0.4437) if MCI-2. Moreover, these fNIRS results were positively correlated with cognitive test scores. These preliminary results showed that fNIRS can be used for distinguishing MCI patients from HCs using mean hemodynamic response as well as temporal features.