How technology is reshaping cognitive assessment: Lessons from the Framingham Heart Study.

Au, Rhoda; Piers, Ryan J.; Devine, Sherral

doi:10.1037/neu0000411

Cited by 47 publications

(34 citation statements)

References 108 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…All subjects were evaluated by trained examiners. For those subjects identified as showing symptoms of cognitive decline, dementia diagnosis was reached by consensus of at least one neurologist and one neuropsychologist; the dementia surveillance and diagnostic procedures have been previously outlined in [5, 21]. The entire dataset for all participants was anonymized prior to analysis.…”

Section: Methodsmentioning

confidence: 99%

An AI-assisted Online Tool for Cognitive Impairment Detection Using Images from the Clock Drawing Test

Amini

Zhang

Hao

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Background: Widespread early dementia detection could drastically increase clinical trial candidates and enable early interventions. Since the Clock Drawing Test (CDT) can be potentially used for diagnosing dementia related diseases, it can be leveraged to devise a computer-aided screening tool. Objective: This work aims to develop an online screening tool by leveraging Artificial Intelligence and the CDT. Methods: Images of an analog clock drawn by 3,263 cognitively intact and 160 impaired subjects were used. First, we processed the images from the CDT by a deep learning algorithm to obtain dementia scores. Then, individuals were classified as belonging to either category by combining CDT image scores with the participants age. Results: We have evaluated the performance of the developed models by applying 5-fold cross validation on 20% of the dataset. The deep learning model generates dementia scores for the CDT images with an Area Under the ROC Curve (AUC) of 81.3%. A composite logistic regression model using age and the generated dementia scores, yielded an average AUC and average weighted F1 score of 92% and 94.4% , respectively. Discussion: CDT images were subjected to distortion consistent with an image drawn on paper and photographed by a cell phone. The model offers a cost-effective and easily deployable mechanism for detecting cognitive impairment online, without the need to visit a clinic.

show abstract

Section: Methodsmentioning

confidence: 99%

An AI-assisted Online Tool for Cognitive Impairment Detection Using Images from the Clock Drawing Test

Amini

Zhang

Hao

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…Selection for dementia review is based on whether participants have shown evidence of cognitive decline, as has been previously described. 4…”

Section: Methodsmentioning

confidence: 99%

Detection of dementia on raw voice recordings using deep learning: A Framingham Heart Study

Xue

Karjadi

Paschalidis

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Background: Identification of reliable, affordable and easy-to-use strategies for detection of dementia are sorely needed. Digital technologies, such as individual voice recordings, offer an attractive modality to assess cognition but methods that could automatically analyze such data without any pre-processing are not readily available. Methods: We used a subset of 1264 digital voice recordings of neuropsychological examinations administered to participants from the Framingham Heart Study (FHS), a community-based longitudinal observational study. The recordings were 73 minutes in duration, on average, and contained at least two speakers (participant and clinician). Of the total voice recordings, 483 were of participants with normal cognition (NC), 451 recordings were of participants with mild cognitive impairment (MCI), and 330 were of participants with dementia. We developed two deep learning models (a two-level long short-term memory (LSTM) network and a convolutional neural network (CNN)), which used the raw audio recordings to classify if the recording included a participant with only NC or only dementia, and also to differentiate between recordings corresponding to non-demented (NC+MCI) and demented participants. Findings: Based on 5-fold cross-validation, the LSTM model achieved a mean (+/-std) area under the sensitivity-specificity curve (AUC) of 0.744+/-0.038, mean accuracy of 0.680+/-0.032, mean sensitivity of 0.719+/-0.112, and mean specificity of 0.652+/-0.089 in predicting cases with dementia from those with normal cognition. The CNN model achieved a mean AUC of 0.805+/-0.027, mean accuracy of 0.740+/-0.033, mean sensitivity of 0.735+/-0.094, and mean specificity of 0.750+/-0.083 in predicting cases with only dementia from those with only NC. For the task related to classification of demented participants from non-demented ones, the LSTM model achieved a mean AUC of 0.659+/-0.043, mean accuracy of 0.701+/-0.057, mean sensitivity of 0.245+/-0.161 and mean specificity of 0.856+/-0.105. The CNN model achieved a mean AUC of 0.730+/-0.039, mean accuracy of 0.735+/-0.046, mean sensitivity of 0.443+/-0.113, and mean specificity of 0.840+/-0.076 in predicting cases with dementia from those who were not demented. Interpretation: This proof-of-concept study demonstrates the potential that raw audio recordings of neuropsychological testing performed on individuals recruited within a community cohort setting can provide a level of screening for dementia.

show abstract

“…Brain aging research began at FHS in 1976, and the detailed cognitive evaluation methods, case ascertainment procedure, and quality control approaches for FHS have been published [11,12]. Brie y, participants were followed-up approximately every two years in the Original cohort and on average every four years in other cohorts.…”

Section: Assessment and Ascertainment Of Cognitive Statusmentioning

confidence: 99%

Severity Distribution of Alzheimer’s Disease Dementia and Mild Cognitive Impairment in the Framingham Heart Study

Yuan

Maserejian

Liu

et al. 2021

JAD

Self Cite

View full text Add to dashboard Cite

Background: Studies providing Alzheimer’s disease (AD) prevalence data have largely neglected to characterize the proportion of AD that is mild, moderate, or severe. Estimates of the severity distribution along the AD continuum, including the mild cognitive impairment (MCI) stage, are important to plan research and allocate future resources, particularly resources targeted at particular stages of disease. Objective: To characterize the distribution of severity of AD dementia and MCI among prevalent cases in the population-based Framingham Heart Study. Methods: Participants (aged 50–94) with prevalent MCI or AD dementia clinical syndrome were cross-sectionally selected from three time-windows of the population-based Framingham Heart Study in 2004-2005 (n = 381), 2006-2007 (n = 422), and 2008-2009 (n = 389). Summary estimates of the severity distribution were achieved by pooling results across time-windows. Diagnosis and severity were assessed by consensus dementia review. MCI-progressive was determined if the participant had documented progression to AD dementia clinical syndrome using longitudinal data. Results: Among AD dementia participants, the pooled percentages were 50.4%for mild, 30.3%for moderate, and 19.3%for severe. Among all MCI and AD participants, the pooled percentages were 29.5%, 19.6%, 25.7%, and 45.2%for MCI-not-progressive, MCI-progressive, mild AD dementia, and the combined group of MCI-progressive and mild AD dementia, respectively. Distributions by age and sex were presented. Conclusion: The finding that half of the people living with AD have mild disease underscores the need for research and interventions to slow decline or prevent progression of this burdensome disease.

show abstract

How technology is reshaping cognitive assessment: Lessons from the Framingham Heart Study.

Cited by 47 publications

References 108 publications

An AI-assisted Online Tool for Cognitive Impairment Detection Using Images from the Clock Drawing Test

An AI-assisted Online Tool for Cognitive Impairment Detection Using Images from the Clock Drawing Test

Detection of dementia on raw voice recordings using deep learning: A Framingham Heart Study

Severity Distribution of Alzheimer’s Disease Dementia and Mild Cognitive Impairment in the Framingham Heart Study

Contact Info

Product

Resources

About