Automatic Speech Classifier for Mild Cognitive Impairment and Early Dementia

The World Health Organization estimates that 50 million people are currently living with dementia worldwide and this figure will almost triple by 2050. Current pharmacological treatments are only symptomatic, and drugs or other therapies are ineffective in slowing down or curing the neurodegenerative process at the basis of dementia. Therefore, early detection of cognitive decline is of the utmost importance to respond significantly and deliver preventive interventions. Recently, the researchers showed that speech alterations might be one of the earliest signs of cognitive defect, observable well in advance before other cognitive deficits become manifest. In this article, we propose a full automated method able to classify the audio file of the subjects according to the progress level of the pathology. In particular, we trained a specific type of artificial neural network, called autoencoder, using the visual representation of the audio signal of the subjects, that is, the spectrogram. Moreover, we used a data augmentation approach to overcome the problem of the large amount of annotated data usually required during the training phase, which represents one of the most major obstacles in deep learning. We evaluated the proposed method using a dataset of 288 audio files from 96 subjects: 48 healthy controls and 48 cognitively impaired participants. The proposed method obtained good classification results compared to the state-of-the-art neuropsychological screening tests and, with an accuracy of 90.57%, outperformed the methods based on manual transcription and annotation of speech.

NCOG-37. THE UTILITY OF NEUROPSYCHOLOGICAL SCREENING IN MULTIDISCIPLINARY CLINICS FOR PEDIATRIC NEURO-ONCOLOGY POPULATIONS

BACKGROUND The purpose of this project was to examine the utility of neuropsychological screening in multidisciplinary clinics for pediatric neuro-oncology populations and determine whether a brief evaluation can adequately screen children for cognitive dysfunction or psychiatric comorbidities. METHODS Children underwent neuropsychological screening as part of their multidisciplinary clinic visit. This screening evaluation consisted of clinical interview, neuropsychological testing, and the completion of emotional/behavioral checklists. Our neuropsychological test battery included the Kaufman Brief Intelligence Test-Second Edition (KBIT-2) and the NIH Toolbox Fluid Cognition Composite (consisting of 6 subtests). RESULTS Over a six-month period 40 children and adolescents with CNS tumors and/or NF1 were seen in our neuropsychology screening clinic. Females made up 37.5% of the sample. The average age was 11.1 years (range=5-25 years). The average time to complete neuropsychological testing was 56.8 minutes. On average, children in this sample met criteria for 1.8 neuropsychological diagnoses with an average of 1.2 new diagnoses that had not been given prior to their neuropsychological screening. The most common new diagnoses were major or mild neurocognitive disorder (mild=27.5%, major=20%), anxiety (27.5%), ADHD (15%), mood disorder (12.5%), and intellectual disability (10%). On average, these children performed lower than expectation on neurocognitive tests, including intellectual functioning (KBIT-2 Verbal IQ=93.1; KBIT-2 Nonverbal IQ=88.7; KBIT-2 Total IQ=90.1), attention (NIH Flanker T-Score=44.1), working memory (NIH Working Memory T-Score=42.9), processing speed (NIH Processing Speed T-Score=39.9), cognitive flexibility (NIH Card Sort T-Score=43.4), memory (NIH Picture Memory T-Score=44.5), and fluid cognition (NIH Fluid Cognition Composite T-Score=38.7). Behavioral health recommendations were needed for 72.5% of this sample. Educational recommendations were needed for 62.5%. CONCLUSIONS Brief neuropsychological screening completed as part of multidisciplinary clinics for pediatric neuro-oncology patients was efficient and useful in identifying neuropsychological diagnoses in children.

A-135 Telehealth Motivationally Enhanced Compensatory Cognitive Training for Older Veterans with Cognitive Complaints: A Pilot Study

Objective This study evaluated the effectiveness of Motivationally Enhanced Compensatory Cognitive Training (ME-CCT), conducted via telehealth. Cognitive performance, self-reported functioning, and activity engagement were expected to improve following participation. Compensatory strategy use was projected to predict posttest memory performance. Method Data from VA electronic medical records was obtained for five veterans (80% male; ages 68–84) who completed an 8-week ME-CCT protocol via telehealth. All completed pre/post neuropsychological screening (RBANS subtests, Oral TMT-B [O-TMT-B]) and self-report measures (PHQ-9, GAD-7, WHODAS, Multifactorial Memory Questionnaire [MMQ]). All had cognitive complaints but no cognitive disorder. Wilcoxon signed-rank test compared pretest-posttest scores. Hierarchical linear regression examined whether strategy use was predictive of posttest memory scores, controlling for age and pretest scores. Results O-TMT-B time (Z = -1.75, p = 0.040, r = −0.78) and errors (Z = -1.73; p = 0.042, r = −0.77) improved significantly, but not semantic fluency (p = 0.358). Immediate (p = 0.055) and Delayed Memory (p = 0.447) did not change. MMQ-Strategy increased (Z = -1.826; p = 0.034, r = −0.82). MMQ-Abilities (p = 0.055, r = −0.82) and MMQ-Satisfaction (p = 0.143 r = −0.48,) did not differ but showed a moderate to large effect. PHQ-9 was lower (Z = -1.841, p = 0.033, r = −0.82); there was a non-significant but large effect on GAD-7 (Z = -1.414, p = 0.079, r = −0.63). WHODAS-Total did not differ (p = 0.231). Use of strategies did not predict Immediate (p = 0.669) or Delayed (p = 0.631) Memory scores. Conclusions Preliminary findings support the use of the ME-CCT protocol with older adults via telehealth, as results suggest possible improvement in aspects of cognition, greater use of strategies, and reductions in mood-related symptoms. Further research with a larger sample is needed to replicate the results and further explore possible outcomes following ME-CCT.

Memory predicted self-care maintenance in patients with heart failure among cognitive functions: Results from a structural equation model

Background: To explain the direction of specific cognitive function training that could prevent potential self-care deficits, the cognitive domain that affects self-care behaviours in patients with heart failure (HF) should be identified. This study aimed to confirm the relationship between major potential variables affecting self-care maintenance behaviours of patients with HF. The study also aimed to determine which cognitive domain predicts self-care maintenance among patients with HF. Methods: This descriptive study involved a secondary analysis using data of 201 patients with HF from two observational studies in three Korean hospitals. The structural equation model using AMOS version 24.0 was constructed to assess the relationships among the variables. The Seoul Neuropsychological Screening Battery was used to assess global cognition, immediate/delayed memory, and executive function, and the Self-Care of Heart Failure Index v.6.2. was used for self-care confidence and maintenance. Results: Self-care maintenance was affected by memory function with a significant direct effect (β=.43, p=.006), as well as self-care confidence (β=.70, p<.001). Memory function and global function indirectly affected self-care maintenance through self-care confidence (β=-.37, p=.002; β=.14, p=.030). Depressive symptoms also had an indirect effect through self-care confidence on self-care maintenance (β=-.21, p=.005). Conclusion: This study confirmed the need for periodic simple screening of the memory function and training to raise it in patients with HF, enhancing their self-care behaviours. It also suggests that supportive intervention is needed to enhance self-confidence as a mediator.

Characteristics of Resting State EEG Power in 80+-Year-Olds of Different Cognitive Status

Compared with healthy older adults, patients with Alzheimer's disease show decreased alpha and beta power as well as increased delta and theta power during resting state electroencephalography (rsEEG). Findings for mild cognitive impairment (MCI), a stage of increased risk of conversion to dementia, are less conclusive. Cognitive status of 213 non-demented high-agers (mean age, 82.5 years) was classified according to a neuropsychological screening and a cognitive test battery. RsEEG was measured with eyes closed and open, and absolute power in delta, theta, alpha, and beta bands were calculated for nine regions. Results indicate no rsEEG power differences between healthy individuals and those with MCI. There were also no differences present between groups in EEG reactivity, the change in power from eyes closed to eyes open, or the topographical pattern of each frequency band. Overall, EEG reactivity was preserved in 80+-year-olds without dementia, and topographical patterns were described for each frequency band. The application of rsEEG power as a marker for the early detection of dementia might be less conclusive for high-agers.

Risks of Obesity in Adolescence: The Role of Physical Activity in Executive Functions

The aim of this study was to analyze the efficiency of a physical activity program on executive performance in obese adolescents. Fifteen adolescents (5 males and 10 females), with a mean age of 14.73 years and an IMC mean of 36.74 participated in the study. None of the participants presented a compulsive eating disorder when screened by the Binge Eating Scale. A pretest and posttest assessment, twelve months later, was conducted by using a neuropsychological battery that evaluated the cognitive flexibility (Comprehensive Trail Making Test), inhibition control (Stroop Neuropsychological Screening Test and by the Frontal Assessment Battery), and planning (Tower of London). ANOVA of repeated measures was performed. The within-subjects tests demonstrated significant statistical differences between the two moments of evaluation at the level of inhibitory control and cognitive flexibility tasks, with higher performances in the second evaluation. These results suggest that the inclusion of obese adolescents in such programs may promote their executive capacities.

Feasibility and Effect of Electroacupuncture on Cognitive Function Domains in Patients with Mild Cognitive Impairment: A Pilot Exploratory Randomized Controlled Trial

Although Electroacupuncture (EA) has been reported to be potentially effective for cognitive disorders, there is limited information about which domains of cognitive function can be improved by EA treatment. Sixty patients with MCI were randomly assigned (1:1:1) to groups to receive 24 sessions over 12 weeks of EA, sham EA, or usual care. In the EA group, electric stimulation was applied at bilateral PC6 and HT7. Various cognitive tests included in the Seoul Neuropsychological Screening Battery II (SNSB-II) were performed at baseline and post-treatment to explore effects of EA on five cognitive domains: attention, language, visuospatial function, memory, and frontal/executive function. Among 60 randomized participants (63.7 ± 7.1 years, 89.7% females), 45 (75%) completed the study. Of the five cognitive function domains of SNSB-II, the T score of visuospatial function showed a tendency to be higher in the EA group than in the usual care group at post-treatment assessment (mean difference: 10.16 (95% CI, 1.14, 19.18), Cohen’s d = 0.72, p = 0.0283). According to the results of this pilot study, the estimated effect size of EA on the visuospatial function of MCI patients compared to usual care was medium. Large-scale clinical trials are needed to confirm effects of EA on cognitive functions.

Short Neuropsychological Assessment in Schizophrenic Patients

Patients with schizophrenia show deficits in executive functions, resembling patterns also found in frontal lesions. The advent of these findings shows the need for appropriate short neuropsychological screening tests in order to assess cognitive function in schizophrenia. Method: We gathered data from a patient group with schizophrenia, and from a control group. We administered 3 short, fast screening tests: MOCA (Montreal Cognitive Assessment), MMSE (Mini Mental State Exam) and FAB (Frontal Assessment Battery). Results: All 3 test were found to be highly correlated with each other, with MOCA being a predictor factor for FAB scores. In measuring the 3 tests’ sensitivity to schizophrenia associated cognitive impairments, we found that FAB was the most sensitive among the three. Conclusions: Our goal was to examine the utility of the FAB test when screening for cognitive deficits in Greek patients with schizophrenia. We found that FAB can be a valuable tool in neuropsychological assessment, but given the varied nature of patients with schizophrenia cognitive profile, it should be accompanied with an extensive battery of tests.