Dataset Information

Scalable diagnostic screening of mild cognitive impairment using AI dialogue agent.

ABSTRACT: The search for early biomarkers of mild cognitive impairment (MCI) has been central to the Alzheimer's Disease (AD) and dementia research community in recent years. To identify MCI status at the earliest possible point, recent studies have shown that linguistic markers such as word choice, utterance and sentence structures can potentially serve as preclinical behavioral markers. Here we present an adaptive dialogue algorithm (an AI-enabled dialogue agent) to identify sequences of questions (a dialogue policy) that distinguish MCI from normal (NL) cognitive status. Our AI agent adapts its questioning strategy based on the user's previous responses to reach an individualized conversational strategy per user. Because the AI agent is adaptive and scales favorably with additional data, our method provides a potential avenue for large-scale preclinical screening of neurocognitive decline as a new digital biomarker, as well as longitudinal tracking of aging patterns in the outpatient setting.

SUBMITTER: Tang F

PROVIDER: S-EPMC7109153 | biostudies-literature | 2020 Mar

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Scalable diagnostic screening of mild cognitive impairment using AI dialogue agent.

Tang Fengyi F Uchendu Ikechukwu I Wang Fei F Dodge Hiroko H HH Zhou Jiayu J

Scientific reports 20200331 1

The search for early biomarkers of mild cognitive impairment (MCI) has been central to the Alzheimer's Disease (AD) and dementia research community in recent years. To identify MCI status at the earliest possible point, recent studies have shown that linguistic markers such as word choice, utterance and sentence structures can potentially serve as preclinical behavioral markers. Here we present an adaptive dialogue algorithm (an AI-enabled dialogue agent) to identify sequences of questions (a di ...[more]

PMID: 32235884

Similar Datasets

Project description:BackgroundArtificial intelligence models tailored to diagnose cognitive impairment have shown excellent results. However, it is unclear whether large linguistic models can rival specialized models by text alone.ObjectiveIn this study, we explored the performance of ChatGPT for primary screening of mild cognitive impairment (MCI) and standardized the design steps and components of the prompts.MethodsWe gathered a total of 174 participants from the DementiaBank screening and classified 70% of them into the training set and 30% of them into the test set. Only text dialogues were kept. Sentences were cleaned using a macro code, followed by a manual check. The prompt consisted of 5 main parts, including character setting, scoring system setting, indicator setting, output setting, and explanatory information setting. Three dimensions of variables from published studies were included: vocabulary (ie, word frequency and word ratio, phrase frequency and phrase ratio, and lexical complexity), syntax and grammar (ie, syntactic complexity and grammatical components), and semantics (ie, semantic density and semantic coherence). We used R 4.3.0. for the analysis of variables and diagnostic indicators.ResultsThree additional indicators related to the severity of MCI were incorporated into the final prompt for the model. These indicators were effective in discriminating between MCI and cognitively normal participants: tip-of-the-tongue phenomenon (P<.001), difficulty with complex ideas (P<.001), and memory issues (P<.001). The final GPT-4 model achieved a sensitivity of 0.8636, a specificity of 0.9487, and an area under the curve of 0.9062 on the training set; on the test set, the sensitivity, specificity, and area under the curve reached 0.7727, 0.8333, and 0.8030, respectively.ConclusionsChatGPT was effective in the primary screening of participants with possible MCI. Improved standardization of prompts by clinicians would also improve the performance of the model. It is important to note that ChatGPT is not a substitute for a clinician making a diagnosis.

Project description:BackgroundMild cognitive impairment (MCI) is often a precursor of dementia, and patients with MCI develop dementia at a higher rate than healthy older adults. Early detection of cognitive decline at the MCI stage supports better planning of care and interventions. At present, the use of virtual reality (VR) in screening for MCI in older adults is promising, but there is little evidence regarding the use of virtual supermarkets to screen for MCI.ObjectiveThe objectives of this study are to validate a VR game-based test, namely, the Virtual Supermarket Program (VSP), for differentiating patients with MCI and healthy controls and to identify cutoff scores for different age levels.MethodsSubjects were recruited from several nursing homes and communities in Changchun, China. They were divided into a healthy control group (n=64) and an MCI group (n=62). All subjects were administered the VSP and a series of neuropsychological examinations. The study determined the optimal cutoff, discriminating validity, concurrent validity, and retest reliability of the VSP. We used the area under the receiver operating characteristic curve (AUC) to evaluate the discriminating validity and obtain the optimal cutoff values. Pearson correlation analysis and the intraclass correlation coefficient were used to evaluate the concurrent validity and retest reliability, respectively.ResultsA cutoff score of 46.4 was optimal for the entire sample, yielding a sensitivity of 85.9% and specificity of 79.0% for differentiating individuals with MCI and healthy controls, and the AUC was 0.870 (95% CI 0.799-0.924). The median index of VSP score was 51.1 (range 42.6-60.0). There was a moderate positive correlation between the VSP total score and Mini-Mental State Examination score (r=0.429, P<.001). There was a strong positive correlation between VSP total score and Montreal Cognitive Assessment score (r=0.645, P<.001). The retest reliability of the VSP was feasible (r=0.588, P=.048).ConclusionsThe VSP is interesting and feasible for subjects. It shows high sensitivity and specificity for the identification of MCI in older adults, which makes it a promising screening method. The VSP may be generalized to older adults in other countries, although some cultural adaptation may be necessary.Trial registrationChinese Clinical Trial Registry ChiCTR2000040074; https://www.chictr.org.cn/showprojen.aspx?proj=64639.

Dataset Information

Scalable diagnostic screening of mild cognitive impairment using AI dialogue agent.

Publications

Scalable diagnostic screening of mild cognitive impairment using AI dialogue agent.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets