| Sumario: | Purpose With the goals of promoting fruitful and healthy lives for the elderly and a society with greater longevity, this study reports the discriminatory performance of a Bayesian-based early detection method of mild cognitive impairment and mild Alzheimer's disease for elderly. Method There are several studies1,2 that suggest associations with Alzheimer's disease and verval performance. This study focuses on cerebral blood flow activation during casual conversation as one of several verbally-based cognitive activities. During the study, an elderly person talks about various topics such as their favourite season, travel, gourmet food, and daily life. With the use of functional near-infrared spectroscopy (fNIRS), that can measure cerebral blood flow activation non-invasively, we collected 42 channels of fNIRS signals from the frontal, right and left temporal areas from 22 elderly participants (7 males and 15 females between the ages of 64 to 89) at a specialized medical institute. The elderly participants were classified into three clinical groups: five patients with mild Alzheimer's disease (AD) and ten participants with mild cognitive impairment (MCI) and seven cognitively normal persons used for controls (CN). The MMSE scores were 29.3±1.0 (CN), 28.6±1.9 (MCI), and 23.4±2.5 (AD). To design an algorithm for computer-aided diagnosis of cognitive impairment in the elderly, we developed a screening process with the help of a specialist in geriatrics. We thus propose a two-phase Bayesian classifier3 (Figure 1) based on the assumption made during the screening process, that firstly checks the suspicion of the cognitive impairment (CI) or not (CN) from given fNIRS signals; if any, and then secondly judges the degree of the impairment for MCI or AD. Results & Discussion We conducted statistical tests using fNIRS signals and examined the detection performance of the proposed Bayesian classifier that can discriminate among elderly individuals with CN, MCI, and AD. Consequently, cross-validation of empirical results indicated this method had an accuracy rate of more than 95% and suggests that the proposed approach may be an effective tool that can be used to screen the elderly for cognitive impairment.
|
|---|