Early detection of cognitive impairment in the elderly based on Bayesian mining using speech prosody and cerebral blood flow activation.

With the aim of providing computer aided diagnosis of dementia, we have developed a non-invasive screening system of the elderly with cognitive impairment. In our previous research, we have studied two data-mining approaches by focusing on speech-prosody and cerebral blood flow (CBF) activation during cognitive tests. On the power of these research results, this paper presents a prosody-CBF hybrid screening system of the elderly with cognitive impairment based on a Bayesian approach. The system is constructed by SPCIR (Speech Prosody-Based Cognitive Impairment Rating) based cutoff as the 1st screening, and, as the 2nd screening, two-phase Bayesian classifier for discriminating among elderly individuals with three clinical groups: elderly individuals with normal cognitive abilities (NC), patients with mild cognitive impairment (MCI), and Alzheimer's disease (AD). This paper also reports the screening examination and discusses the cost-effectiveness and the discrimination performance of the proposed system for early detection of cognitive impairment in elderly subjects.