Automatic Detection of Cognitive Impairments through Acoustic Analysis of Speech

Background: Early detection of mild cognitive impairment is crucial in the prevention of Alzheimer’s disease. The aim of the present study was to identify whether acoustic features can help differentiate older, independent community-dwelling individuals with cognitive impairment from healthy controls. Methods: A total of 8779 participants (mean age 74.2 ± 5.7 in the range of 65-96, 3907 males and 4872 females) with different cognitive profiles, namely healthy controls, mild cognitive impairment, global cognitive impairment (defined as a Mini Mental State Examination score of 20-23), and mild cognitive impairment with global cognitive impairment (a combined status of mild cognitive impairment and global cognitive impairment), were evaluated in short-sentence reading tasks, and their acoustic features, including temporal features (such as duration of utterance, number and length of pauses) and spectral features (F0, F1, and F2), were used to build a machine learning model to predict their cognitive impairments. Results: The classification metrics from the healthy controls were evaluated through the area under the receiver operating characteristic curve and were found to be 0.61, 0.67, and 0.77 for mild cognitive impairment, global cognitive impairment, and mild cognitive impairment with global cognitive impairment, respectively. Conclusion: Our machine learning model revealed that individuals’ acoustic features can be employed to discriminate between healthy controls and those with mild cognitive impairment with global cognitive impairment, which is a more severe form of cognitive impairment compared with mild cognitive impairment or global cognitive impairment alone. It is suggested that language impairment increases in severity with cognitive impairment.

Download Full-text

An SBM-based machine learning model for identifying mild cognitive impairment in patients with Parkinson's disease

Journal of the Neurological Sciences ◽

10.1016/j.jns.2020.117077 ◽

2020 ◽

Vol 418 ◽

pp. 117077

Author(s):

Jiahui Zhang ◽

You Li ◽

Yuyuan Gao ◽

Jinlong Hu ◽

Biao Huang ◽

...

Keyword(s):

Machine Learning ◽

Parkinson’S Disease ◽

Parkinson's Disease ◽

Cognitive Impairment ◽

Mild Cognitive Impairment ◽

Learning Model ◽

Machine Learning Model

Download Full-text

MRI-Based Radiomic Machine-Learning Model May Accurately Distinguish between Subjects with Internet Gaming Disorder and Healthy Controls

Brain Sciences ◽

10.3390/brainsci12010044 ◽

2021 ◽

Vol 12 (1) ◽

pp. 44

Author(s):

Xu Han ◽

Lei Wei ◽

Yawen Sun ◽

Ying Hu ◽

Yao Wang ◽

...

Keyword(s):

Machine Learning ◽

Diffusion Tensor ◽

Internet Gaming Disorder ◽

Learning Model ◽

Brain Morphology ◽

Healthy Controls ◽

Gaming Disorder ◽

Internet Gaming ◽

Machine Learning Model ◽

And Diffusion

Purpose To identify cerebral radiomic features related to the diagnosis of Internet gaming disorder (IGD) and construct a radiomics-based machine-learning model for IGD diagnosis. Methods A total of 59 treatment-naïve subjects with IGD and 69 age- and sex-matched healthy controls (HCs) were recruited and underwent anatomic and diffusion-tensor magnetic resonance imaging (MRI). The features of the morphometric properties of gray matter and diffusion properties of white matter were extracted for each participant. After excluding the noise feature with single-factor analysis of variance, the remaining 179 features were included in an all-relevant feature selection procedure within cross-validation loops to identify features with significant discriminative power. Random forest classifiers were constructed and evaluated based on the identified features. Results No overall differences in the total brain volume (1,555,295.64 ± 152316.31 mm3 vs. 154,491.19 ± 151,241.11 mm3), total gray (709,119.83 ± 59,534.46 mm3 vs. 751,018.21 ± 58,611.32 mm3) and white (465,054.49 ± 51,862.65 mm3 vs. 470,600.22 ± 47,006.67 mm3) matter volumes, and subcortical region volume (63,882.71 ± 5110.42 mm3 vs. 64,764.36 ± 4332.33 mm3) between the IGD and HC groups were observed. The mean classification accuracy was 73%. An altered cortical shape in the bilateral fusiform, left rostral middle frontal (rMFG), left cuneus, left parsopercularis (IFG), and regions around the right uncinate fasciculus (UF) and left internal capsule (IC) contributed significantly to group discrimination. Conclusions: Our study found the brain morphology alterations between IGD subjects and HCs through a radiomics-based machine-learning method, which may help revealing underlying IGD-related neurobiology mechanisms.

Download Full-text

Similar Theory of Mind Deficits in Community Dwelling Older Adults with Vascular Risk Profile and Patients with Mild Cognitive Impairment: The Case of Paradoxical Sarcasm Comprehension

Brain Sciences ◽

10.3390/brainsci11050627 ◽

2021 ◽

Vol 11 (5) ◽

pp. 627

Author(s):

Glykeria Tsentidou ◽

Despina Moraitou ◽

Magda Tsolaki

Keyword(s):

Risk Factors ◽

Older Adults ◽

Cognitive Impairment ◽

Mild Cognitive Impairment ◽

Theory Of Mind ◽

Vascular Risk Factors ◽

Community Dwelling ◽

Healthy Controls ◽

Vascular Risk ◽

Prediction And Prevention

Recent studies deal with disorders and deficits caused by vascular syndrome in efforts for prediction and prevention. Cardiovascular health declines with age due to vascular risk factors, and this leads to an increasing risk of cognitive decline. Mild cognitive impairment (MCI) is defined as the negative cognitive changes beyond what is expected in normal aging. The purpose of the study was to compare older adults with vascular risk factors (VRF), MCI patients, and healthy controls (HC) in social cognition and especially in theory of mind ability (ToM). The sample comprised a total of 109 adults, aged 50 to 85 years (M = 66.09, SD = 9.02). They were divided into three groups: (a) older adults with VRF, (b) MCI patients, and (c) healthy controls (HC). VRF and MCI did not differ significantly in age, educational level or gender as was the case with HC. Specifically, for assessing ToM, a social inference test was used, which was designed to measure sarcasm comprehension. Results showed that the performance of the VRF group and MCI patients is not differentiated, while HC performed higher compared to the other two groups. The findings may imply that the development of a vascular disorder affecting vessels of the brain is associated from its “first steps” to ToM decline, at least regarding specific aspects of it, such as paradoxical sarcasm understanding.

Download Full-text

Plasma d-glutamate levels for detecting mild cognitive impairment and Alzheimer’s disease: Machine learning approaches

Journal of Psychopharmacology ◽

10.1177/0269881120972331 ◽

2021 ◽

Vol 35 (3) ◽

pp. 265-272 ◽

Cited By ~ 1

Author(s):

Chun-Hung Chang ◽

Chieh-Hsin Lin ◽

Chieh-Yu Liu ◽

Chih-Sheng Huang ◽

Shaw-Ji Chen ◽

...

Keyword(s):

Machine Learning ◽

Cognitive Impairment ◽

Mild Cognitive Impairment ◽

Random Forest ◽

Naive Bayes ◽

Mmse Score ◽

Learning Algorithms ◽

Machine Learning Algorithms ◽

Healthy Controls ◽

Peripheral Plasma

Background: d-glutamate, which is involved in N-methyl-d-aspartate receptor modulation, may be associated with cognitive ageing. Aims: This study aimed to use peripheral plasma d-glutamate levels to differentiate patients with mild cognitive impairment (MCI) and Alzheimer’s disease (AD) from healthy individuals and to evaluate its prediction ability using machine learning. Methods: Overall, 31 healthy controls, 21 patients with MCI and 133 patients with AD were recruited. Serum d-glutamate levels were measured using high-performance liquid chromatography (HPLC). Cognitive deficit severity was assessed using the Clinical Dementia Rating scale and the Mini-Mental Status Examination (MMSE). We employed four machine learning algorithms (support vector machine, logistic regression, random forest and naïve Bayes) to build an optimal predictive model to distinguish patients with MCI or AD from healthy controls. Results: The MCI and AD groups had lower plasma d-glutamate levels (1097.79 ± 283.99 and 785.10 ± 720.06 ng/mL, respectively) compared to healthy controls (1620.08 ± 548.80 ng/mL). The naïve Bayes model and random forest model appeared to be the best models for determining MCI and AD susceptibility, respectively (area under the receiver operating characteristic curve: 0.8207 and 0.7900; sensitivity: 0.8438 and 0.6997; and specificity: 0.8158 and 0.9188, respectively). The total MMSE score was positively correlated with d-glutamate levels ( r = 0.368, p < 0.001). Multivariate regression analysis indicated that d-glutamate levels were significantly associated with the total MMSE score ( B = 0.003, 95% confidence interval 0.002–0.005, p < 0.001). Conclusions: Peripheral plasma d-glutamate levels were associated with cognitive impairment and may therefore be a suitable peripheral biomarker for detecting MCI and AD. Rapid and cost-effective HPLC for biomarkers and machine learning algorithms may assist physicians in diagnosing MCI and AD in outpatient clinics.

Download Full-text

Feasibility of In-Home Sensor Monitoring to Detect Mild Cognitive Impairment in Aging Military Veterans: Prospective Observational Study

JMIR Formative Research ◽

10.2196/16371 ◽

2020 ◽

Vol 4 (6) ◽

pp. e16371

Author(s):

Adriana Seelye ◽

Mira Isabelle Leese ◽

Katherine Dorociak ◽

Nicole Bouranis ◽

Nora Mattek ◽

...

Keyword(s):

Cognitive Impairment ◽

Mild Cognitive Impairment ◽

Older Adult ◽

Scale Score ◽

Rating Scale ◽

Military Veterans ◽

Community Dwelling ◽

Daily Functioning ◽

Healthy Controls ◽

Military Veteran

Background Aging military veterans are an important and growing population who are at an elevated risk for developing mild cognitive impairment (MCI) and Alzheimer dementia, which emerge insidiously and progress gradually. Traditional clinic-based assessments are administered infrequently, making these visits less ideal to capture the earliest signals of cognitive and daily functioning decline in older adults. Objective This study aimed to evaluate the feasibility of a novel ecologically valid assessment approach that integrates passive in-home and mobile technologies to assess instrumental activities of daily living (IADLs) that are not well captured by clinic-based assessment methods in an aging military veteran sample. Methods Participants included 30 community-dwelling military veterans, classified as healthy controls (mean age 72.8, SD 4.9 years; n=15) or MCI (mean age 74.3, SD 6.0 years; n=15) using the Clinical Dementia Rating Scale. Participants were in relatively good health (mean modified Cumulative Illness Rating Scale score 23.1, SD 2.9) without evidence of depression (mean Geriatrics Depression Scale score 1.3, SD 1.6) or anxiety (mean generalized anxiety disorder questionnaire 1.3, SD 1.3) on self-report measures. Participants were clinically assessed at baseline and 12 months later with health and daily function questionnaires and neuropsychological testing. Daily computer use, medication taking, and physical activity and sleep data were collected via passive computer monitoring software, an instrumented pillbox, and a fitness tracker watch in participants’ environments for 12 months between clinical study visits. Results Enrollment began in October 2018 and continued until the study groups were filled in January 2019. A total of 201 people called to participate following public posting and focused mailings. Most common exclusionary criteria included nonveteran status 11.4% (23/201), living too far from the study site 9.4% (19/201), and having exclusionary health concerns 17.9% (36/201). Five people have withdrawn from the study: 2 with unanticipated health conditions, 2 living in a vacation home for more than half of the year, and 1 who saw no direct benefit from the research study. At baseline, MCI participants had lower Montreal Cognitive Assessment (P<.001) and higher Functional Activities Questionnaire (P=.04) scores than healthy controls. Over seven months, research personnel visited participants’ homes a total of 73 times for technology maintenance. Technology maintenance visits were more prevalent for MCI participants (P=.04) than healthy controls. Conclusions Installation and longitudinal deployment of a passive in-home IADL monitoring platform with an older adult military veteran sample was feasible. Knowledge gained from this pilot study will be used to help develop acceptable and effective home-based assessment tools that can be used to passively monitor cognition and daily functioning in older adult samples.

Download Full-text

Neuro-anatomical and neuro-molecular markers in salience network and their integration in machine learning-based prediction of cognitive dysfunction in mild cognitive impairment

10.1101/2021.12.22.473775 ◽

2021 ◽

Author(s):

Ganesh Chand ◽

Deepa S. Thakuri ◽

Bhavin Soni

Keyword(s):

Machine Learning ◽

Cognitive Impairment ◽

Molecular Markers ◽

Mild Cognitive Impairment ◽

Molecular Mechanisms ◽

Anterior Cingulate ◽

Salience Network ◽

Healthy Controls ◽

Positron Emission ◽

Functional Mechanisms

Recent studies indicate disrupted functional mechanisms of salience network regions, especially right anterior insula (RAI), left AI (LAI), and anterior cingulate cortex (ACC), in mild cognitive impairment (MCI). However, the underlying neuro-anatomical and neuro-molecular mechanisms in these regions are not clearly understood yet. It is also unknown whether integration of multi-modal neuro-anatomical and neuro-molecular markers could predict cognitive impairment better in MCI. Herein we quantified neuro-anatomical volumetric markers via structural magnetic resonance imaging (sMRI) and neuro-molecular amyloid markers via positron emission tomography with Pittsburgh compound B (PET PiB) in SN regions of MCI (n = 33) and healthy controls (n = 27). We found that neuro-anatomical markers are significantly reduced while neuro-molecular markers are significantly elevated in SN nodes of MCI compared to healthy controls (p < 0.05). These altered markers in MCI patients were associated with their worse cognitive performance (p < 0.05). Our machine learning-based modeling further suggested that the integration of multi-modal markers predicts cognitive impairment in MCI superiorly compared to using single modality-specific markers. Overall, these findings shed light on the underlying neuro-anatomical volumetric and neuro-molecular amyloid alterations in SN regions and show the significance of multi-modal markers integration approach in better predicting cognitive impairment in MCI.

Download Full-text

Feasibility of In-Home Sensor Monitoring to Detect Mild Cognitive Impairment in Aging Military Veterans: Prospective Observational Study (Preprint)

10.2196/preprints.16371 ◽

2019 ◽

Author(s):

Adriana Seelye ◽

Mira Isabelle Leese ◽

Katherine Dorociak ◽

Nicole Bouranis ◽

Nora Mattek ◽

...

Keyword(s):

Cognitive Impairment ◽

Mild Cognitive Impairment ◽

Older Adult ◽

Scale Score ◽

Rating Scale ◽

Military Veterans ◽

Community Dwelling ◽

Daily Functioning ◽

Healthy Controls ◽

Military Veteran

BACKGROUND Aging military veterans are an important and growing population who are at an elevated risk for developing mild cognitive impairment (MCI) and Alzheimer dementia, which emerge insidiously and progress gradually. Traditional clinic-based assessments are administered infrequently, making these visits less ideal to capture the earliest signals of cognitive and daily functioning decline in older adults. OBJECTIVE This study aimed to evaluate the feasibility of a novel ecologically valid assessment approach that integrates passive in-home and mobile technologies to assess instrumental activities of daily living (IADLs) that are not well captured by clinic-based assessment methods in an aging military veteran sample. METHODS Participants included 30 community-dwelling military veterans, classified as healthy controls (mean age 72.8, SD 4.9 years; n=15) or MCI (mean age 74.3, SD 6.0 years; n=15) using the Clinical Dementia Rating Scale. Participants were in relatively good health (mean modified Cumulative Illness Rating Scale score 23.1, SD 2.9) without evidence of depression (mean Geriatrics Depression Scale score 1.3, SD 1.6) or anxiety (mean generalized anxiety disorder questionnaire 1.3, SD 1.3) on self-report measures. Participants were clinically assessed at baseline and 12 months later with health and daily function questionnaires and neuropsychological testing. Daily computer use, medication taking, and physical activity and sleep data were collected via passive computer monitoring software, an instrumented pillbox, and a fitness tracker watch in participants’ environments for 12 months between clinical study visits. RESULTS Enrollment began in October 2018 and continued until the study groups were filled in January 2019. A total of 201 people called to participate following public posting and focused mailings. Most common exclusionary criteria included nonveteran status 11.4% (23/201), living too far from the study site 9.4% (19/201), and having exclusionary health concerns 17.9% (36/201). Five people have withdrawn from the study: 2 with unanticipated health conditions, 2 living in a vacation home for more than half of the year, and 1 who saw no direct benefit from the research study. At baseline, MCI participants had lower Montreal Cognitive Assessment (P<.001) and higher Functional Activities Questionnaire (P=.04) scores than healthy controls. Over seven months, research personnel visited participants’ homes a total of 73 times for technology maintenance. Technology maintenance visits were more prevalent for MCI participants (P=.04) than healthy controls. CONCLUSIONS Installation and longitudinal deployment of a passive in-home IADL monitoring platform with an older adult military veteran sample was feasible. Knowledge gained from this pilot study will be used to help develop acceptable and effective home-based assessment tools that can be used to passively monitor cognition and daily functioning in older adult samples.

Download Full-text