An internet-based screening test for Mild Cognitive Impairment (MCI) from spontaneous speech

Abstract Background: Even today there is a great diagnostic uncertainty in the prodromal stages of Alzheimer’s disease (AD). Our recent effort has focused on the earliest detectable indicators of cognitive decline in mild cognitive impairment (MCI). Since language impairment has been demonstrated even in the mild stage of AD (1), the aim of the present study is to develop a sensitive neuropsychological screening method on the basis of spontaneous speech production in a memory task, which could be suitable for recognizing MCI via an internet based interactive computerized test and software in the future. Methods: Participants of the study: 25 normal controls (NC), and 25 clinically diagnosed MCI patients. Spontaneous speech was generated in two ways, first, through visual input -- recalling the content of 2 short black and white films (one direct, one delayed) -- and second, answering one question. Phonetical parameters such as speech tempo, hesitation ratio, length and number of silent and filled pauses, length of utterance were measured in the data (using PRAAT). Clinical tests were: Clock Drawing Test (2), Mini-Mental State Examination (3), ADAS-Cog (4), Early Mental Test (EMT) (5). Results: Statistically significant differences were found in speech tempo in the delayed recall task, as well as in the length of silent pauses during the question-answering spontaneous speech task. The phonetical analysis of the spontaneous speech showed differences between the examined groups in articulation rate, speech tempo, silent pause, hesitation ratio, length of utterance and break-to-utterance ratio. Conclusion: The higher hesitation ratio of the MCI group might reflect increased recall time. It is likely that the phonetical analysis of spontaneous speech in our design can be fruitfully exploited in implementing a new, sensitive web based tool for screening MCI for the community