Research Of The Different Performance Between Different Cognitive Impairment Patients In Graphs Analysis Of The Animal Verbal Fluency Test

Background and objectiveCognitive impairment, usually refers to different degree of cognitive damage caused by a variety of reasons, according to the different damage degree, it can be divided into mild cognitive impairment (MCI) and Dementia. Mild cognitive impairment (MCI) is a disease which present memory damage and (or) other domain mild impairment, but the ability of daily living is unaffected and the damage degree have not reached the dementia. Alzheimer’s disease which is associated with age, is the most common type of dementia, it is characterized by memory loss and cognitive impairment progressively, its symptom in the early stage is memory impairment, especially it is difficult to remember the recent events (short-term memory loss).Verbal Fluency Test (VFT) is a simple neuropsychological test which involved with the cognitive function like memory function, language function (comprehend, semantic knowledge, named ability), executive function. Animal Fluency Test (AFT) is one simple and easy method of VFT for semantic fluency examination, in which patients are required to say animal names within one minute as more as possible. Although AFT is used widely in neuropsychological assessment at present, the standard measures for its analysis and evaluation is limited to the number of correct words produced, without considering other clinically relevant information in the test, such as the count of error words, the count and fraction of repetitive words.In this research, we make AFT speech graph via the software SpeechGraphs created by the brain institute of federal university in Rio grande do norte (Natal, Brazil), and describe the different characteristics of groups by evaluating the AFT related data and 13 speech graph attributes (SGA). This study is intended to analyze the differences of performance and evaluate the diagnostic values between MCI and AD in SpeechGraphs of the AFT.Materials and methodsThe patients who had undertaken neuropsychological assessment at cognitive function testing room from July 2015 to March 2016 in Department of Neurology, Qilu Hospital of Shandong University, were screened and divided into three groups according to dementia subtypes and severity. (I)"probable MCI" group (MCI), "probable AD" group(AD), which contains two subtype:"probable mild AD " group (MMSE>20 points), "probable moderate to severe AD "group (MMSE≤20 points),and the normal control group (Con group) in the same time period.The participants performed the AFT, category of animals, in which they were asked to produce the maximum names of animals within one minute, and explicit/implicit instructions were given to avoid repetitions. All the words were recorded, including repeated and error words. The scoring procedure included:total of words produced, total of correct words, total of errors words, total of repeated words, and the percentage of repeated words produced by each participant. The scores in this task were not taken into account in the diagnosis adjudication of each participant.Make AFT speech graph via the software SpeechGraphs created by the brain institute of federal university in Rio grande do norte (Natal, Brazil), and describe the different characteristics of groups by evaluating the AFT related data and 13 speech graph attributes (SGA).SPSS 17.0 was used to perform statistical analyses. The differences in age, education, MMSE score, AFT score of the three groups were compared through the Kruskal-Wallis test. We used the Kruskal-Wallis test to analyze the differences of the three independent groups in AFTand SGA, and the Wilcoxon Rank sum test for two independent samples in Con group, MCI group, AD group and Con group, MCI group, mild AD group, moderate to severe AD group. Bonferroni correction was applied to all analyses. Draw the receiver-operating characteristic curve (receiver operating characteristic curve, the ROC) and to evaluate diagnostic value of SGA based on the sensitivity, specificity and the area under the receiver operating characteristic curve (AUC).Results30 patients were selected into "probable MCI" group; 63 patients were selected into "probable AD" group, in which 30 patients were selected into "probable mild AD" group; 33 patients were selected into "probable moderate to severe AD" group; 30 subjects were selected into control group. There was no significant difference in age between each group, but the age of AD group was larger than other two groups. In the aspect of educational level, MMSE score, Con group showed statistically significant differences with other two groups. The educational level of Con group was obviously higher than other two groups. In the aspect of AFT score and ADL, Con group showed statistically significant differences with AD group, but no differences with MCI group.The groups significantly differed in the performance on number of total words and produced correct animal names, and in the SGA of nodes, edges, diameter, average short path and density. Con group produced more nodes, a network with larger diameter and less dense, when compared with the MCI and AD groups.While the AD group was divided into mild AD group and moderate to serve AD group, the pairwise comparisons of the four groups showed, compared MCI group, mild AD group and moderate to severe AD group with Con group, showed differences in the AFT.TT, AFT.C, N, E, Density, Diameter, ASP. Compared the MCI group, mild AD group with moderate to serve AD group, differences in AFT.TT, AFT.C, N, E, Density, Diameter, ASP was obvious, but MCI group and mild AD group is similar in those.Using ROC curve, the sensitivity and specificity of SGA for diagnosis of MCI was80.0% and 83.3%(AUC=0.743), the sensitivity and specificity of SGA for diagnosis of mild AD is 80.0% and70.0%(AUC=0.706), and the sensitivity and specificity of SGA for diagnosis of moderate to severe AD is93.3% and 51.5%(AUC =0.677).Conclusion(1) Compared with the normal elderly, AFT analysis in cognitive impairment patients, produce the total number of words and animals names less, with high repeatability.(2) With the cognitive function decline of cognitive impairment patients, the more serious stage of cognitive impairment, the AFT speech graph may have less nodes and edges, larger density, shorter diameter and average shortest path.(3) SGA can be used to distinguish Con, MCI, mild AD and moderate to severe AD with high sensitivity, and is of certain value in the diagnosis of cognitive impairment, but the specificity is low, and it can’t be a diagnostic tool independently, still need to unite with other tools.