The current state of artificial intelligence-augmented digitized neurocognitive screening test

Sirilertmekasakul, Chananchida; Rattanawong, Wanakorn; Gongvatana, Assawin; Srikiatkhachorn, Anan

doi:10.3389/fnhum.2023.1133632

Cited by 6 publications

(4 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Now, with the advancement in AI-based translators, such tools can offer a cost-effective solution to any language barrier when delivering care to populations with high language diversity (see example in Panayiotou et al 94 ). There is, for instance, an interest in the development of AI-based cognitive tests, 95 including chatbots powered by the like of ChatGPT to conduct cognitive tests, but these chatbots based on large language models do not perform well in all languages (e.g., Seghier 96 ). We reckon that the use of AI in neuropsychology is still in its infancy, facing many limitations with respect to generalizability of AI algorithms and availability of high-quality and bias-free data.…”

Section: Implications For the Global Fight Against Dementiamentioning

confidence: 99%

“… 94 ). There is, for instance, an interest in the development of AI‐based cognitive tests, 95 including chatbots powered by the like of ChatGPT to conduct cognitive tests, but these chatbots based on large language models do not perform well in all languages (e.g., Seghier 96 ). We reckon that the use of AI in neuropsychology is still in its infancy, facing many limitations with respect to generalizability of AI algorithms and availability of high‐quality and bias‐free data.…”

Section: Implications For the Global Fight Against Dementiamentioning

confidence: 99%

See 1 more Smart Citation

Cognitive decline assessment in speakers of understudied languages

Hatahet,

Roser,

Seghier

2023

A&D Transl Res & Clin Interv

View full text Add to dashboard Cite

Projected trends in population aging have forecasted a massive increase in the number of people with dementia, in particular in sub‐Saharan Africa and the Middle East and North Africa (MENA) region. Cognitive decline is a significant marker for dementia, typically assessed with standardized neuropsychological tools that have been validated in some well‐researched languages such as English. However, with the existing language diversity, current tools cannot cater to speakers of understudied languages, putting these populations at a disadvantage when it comes to access to early and accurate diagnosis of dementia. Here, we shed light on the detrimental impact of this language gap in the context of the MENA region, highlighting inadequate tools and an unacceptable lack of expertise for a MENA population of a half billion people. Our perspective calls for more research to unravel the exact impact of the language gap on the quality of cognitive decline assessment in speakers of understudied languages.Highlights Cognitive decline is a marker for dementia, assessed with neuropsychological tests. There is a lack of culturally valid tests for speakers of understudied languages. For example, suboptimal cognitive tests are used in the Middle East and North Africa region. Linguistic diversity should be considered in the development of cognitive tests.

show abstract

Section: Implications For the Global Fight Against Dementiamentioning

confidence: 99%

Section: Implications For the Global Fight Against Dementiamentioning

confidence: 99%

Cognitive decline assessment in speakers of understudied languages

Hatahet,

Roser,

Seghier

2023

A&D Transl Res & Clin Interv

View full text Add to dashboard Cite

show abstract

“…10 AI has emerged as a promising tool in healthcare, particularly in the area of cognitive impairment diagnosis. 11,12 AI can provide a more accurate and standardized process for diagnosing and predicting disease. 13 Thabtah et al conducted a comprehensive analysis of mobile-based AD and MCI screening apps, highlighting the potential of AI applications for early detection and diagnosis of dementia and exploring AI to improve access to healthcare services.…”

Section: Introductionmentioning

confidence: 99%

Text dialogue analysis Based ChatGPT for Primary Screening of Mild Cognitive Impairment

Wang

Liu

et al. 2023

Preprint

View full text Add to dashboard Cite

Background: AI 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. Objectives: We would explore the effectiveness of ChatGPT for primary screening of mild cognitive impairment (MCI) and standardize the design steps and components of the prompt. Methods: We obtained 174 participants from the DementiaBank screening and classified 70% of them into the training set and 30% into the test set. Three dimensions of variables were incorporated, including: vocabulary, syntax and grammar, and semantics. These variables were generated from published studies and statistical analyses. We used R 4.3.0. for the analysis of variables and diagnostic indicators. Results: The final variables included by published studies included: word frequency and word ratio, phrase frequency and phrase ratio, lexical complexity, syntactic complexity, grammatical components, semantic density, and semantic coherence; variables included in the analysis included: tip-of-the-tongue phenomenon (P < 0.001), difficulty with complex ideas (P < 0.001), and memory issues (P < 0.001). The final GPT4 model achieved the sensitivity (SEN) of 0.8636, specificity (SPE) of 0.9487 and area under the curve (AUC) of 0.9062 on the training set; on the test set, the SEN, SPE and AUC reached 0.7727, 0.8333 and 0.8030, respectively. The prompt consisted of five main parts, including character setting, scoring system setting, indicator setting, output setting, and explanatory information setting. Conclusion: ChatGPT was effective in primary screening of participants with possible MCI. Improved standardization of prompts by professional 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.

show abstract

“…AI has emerged as a promising tool in health care, particularly in cognitive impairment diagnosis [11,12]. AI can provide a more accurate and standardized process for diagnosing and predicting diseases [13].…”

Section: Introductionmentioning

confidence: 99%

Text Dialogue Analysis for Primary Screening of Mild Cognitive Impairment: Development and Validation Study

Wang,

Liu,

et al. 2023

J Med Internet Res

View full text Add to dashboard Cite

Background Artificial 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. Objective In 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. Methods We 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. Results Three 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. Conclusions ChatGPT 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.

show abstract

The current state of artificial intelligence-augmented digitized neurocognitive screening test

Cited by 6 publications

References 37 publications

Cognitive decline assessment in speakers of understudied languages

Cognitive decline assessment in speakers of understudied languages

Text dialogue analysis Based ChatGPT for Primary Screening of Mild Cognitive Impairment

Text Dialogue Analysis for Primary Screening of Mild Cognitive Impairment: Development and Validation Study

Contact Info

Product

Resources

About