Alyssa N. Kaser
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Alyssa N. Kaser has a total of 49 co-authors. Here are the top co-authors by number of shared publications:
- Laura H. Lacritz
Department of Psychiatry, The University of Texas Southwestern Medical Center
4 shared publications - Oscar R. Kronenberger
University of Texas Southwestern Medical Center
3 shared publications - Andrew M. Kiselica
Institute of Gerontology, University of Georgia
2 shared publications - Colin Munro Cullum
Department of Psychiatry, University of Texas Southwestern Medical Center
2 shared publications - Jeff Schaffert
Department of Psychiatry, University of Texas Southwestern Medical Center
2 shared publications Alberto J Coca
AcceXible Impacto, Sociedad Limitada, Bilbao, Spain
1 shared publication
Latest Publications
- Open Access
Feasibility of remote self-administered cognitive assessment in an Alzheimer’s disease treatment clinic
Introduction: Remote cognitive assessments offer an accessible, scalable way to support clinical characterization and longitudinal monitoring of cognition. This study evaluates the feasibility of implementing Mayo Test Drive (MTD), a remote, self-administered digital cognitive assessment, within a specialized clinic focused on eligibility assessments for anti-amyloid therapies for Alzheimer’s disease (AD). Method: Two-hundred and nine ( N =209) patients (M age =71.2, 53% female) were evaluated through the Mayo Clinic Rochester Alzheimer’s Disease Treatment Clinic (ADTC) and had the opportunity to complete MTD, either as a remote, patient-initiated session through a patient-facing healthcare portal or during an in-clinic neuropsychological evaluation. Results: Clinical diagnoses included 55.2% mild cognitive impairment (MCI), 31.0% mild dementia, and 11.6% moderate dementia, in addition to 2.2% cognitively unimpaired (CU), as determined by neurology consensus at a multidisciplinary case conference. Among those who initiated MTD ( n =197/209), 97.5% completed a session (70.6% remotely, 26.9% in clinic). Completion rates were high across diagnostic groups and did not significantly differ by clinical severity ( p =.12): CU (100%), MCI (99.1%), mild dementia (97.0%), and moderate dementia (89.5%). Remote completion rates overall were high (97.9%). Among completers, administration setting (remote vs. in-clinic) did not significantly differ by clinical stage, χ²(3)=2.37, p =.49. Reasons for non-initiation included not receiving or opening portal messages and other/unknown factors. Discussion: Remote digital cognitive assessment is feasible within the context of clinic eligibility evaluations for AD therapies. The high completion rates of MTD across clinical impairment levels in this setting support development of paths for broader use in cognitive characterization in AD care.
Open AccessFeasibility of remote self-administered cognitive assessment in an Alzheimer’s disease treatment clinic
Introduction: Remote cognitive assessments offer an accessible, scalable way to support clinical characterization and longitudinal monitoring of cognition. This study evaluates the feasibility of implementing Mayo Test Drive (MTD), a remote, self-administered digital cognitive assessment, within a specialized clinic focused on eligibility assessments for anti-amyloid therapies for Alzheimer’s disease (AD). Method: Two-hundred and nine ( N =209) patients (M age =71.2, 53% female) were evaluated through the Mayo Clinic Rochester Alzheimer’s Disease Treatment Clinic (ADTC) and had the opportunity to complete MTD, either as a remote, patient-initiated session through a patient-facing healthcare portal or during an in-clinic neuropsychological evaluation. Results: Clinical diagnoses included 55.2% mild cognitive impairment (MCI), 31.0% mild dementia, and 11.6% moderate dementia, in addition to 2.2% cognitively unimpaired (CU), as determined by neurology consensus at a multidisciplinary case conference. Among those who initiated MTD ( n =197/209), 97.5% completed a session (70.6% remotely, 26.9% in clinic). Completion rates were high across diagnostic groups and did not significantly differ by clinical severity ( p =.12): CU (100%), MCI (99.1%), mild dementia (97.0%), and moderate dementia (89.5%). Remote completion rates overall were high (97.9%). Among completers, administration setting (remote vs. in-clinic) did not significantly differ by clinical stage, χ²(3)=2.37, p =.49. Reasons for non-initiation included not receiving or opening portal messages and other/unknown factors. Discussion: Remote digital cognitive assessment is feasible within the context of clinic eligibility evaluations for AD therapies. The high completion rates of MTD across clinical impairment levels in this setting support development of paths for broader use in cognitive characterization in AD care.
- 5 February 2026
- Published by Oxford University Press (OUP) in Archives of Clinical Neuropsychology
- Vol. 41 (2)
Texas Card Sorting Test: a Brief Test of Executive Functioning with Age-Adjusted Norms and External Validation
Objective: The Texas Card Sorting Test (TCST) is a brief test of executive functioning that involves sorting cards into groups based on like dimensions. The TCST is shorter than most available executive function measures, requires minimal language demands, and excludes negative feedback. This study generated age-adjusted normative data for the TCST in a healthy aging, normal control (NC) sample, examined its sensitivity and specificity, assessed convergent validity, and compared NC performance to an Alzheimer’s disease clinical syndrome (ADCS) sample. Method: TCST data from 164 NC [Mage = 69.60 (SD = 8.6); Meducation = 15.46 years (SD = 2.83)] and 92 ADCS participants [Mage = 77.73 (SD = 7.55); Meducation = 13.65 years (SD = 3.53)] were obtained as part of a larger neuropsychological battery at Texas Alzheimer’s Research and Care Consortium (TARCC) sites from which age-adjusted normative data were derived and sensitivity, specificity, diagnostic accuracy, and validity were examined. Results: TCST Logical Sorts (LS), or correct card sorts, were significantly correlated with age, education, and other executive function measures. Lower but significant correlations were seen with attention and memory measures. TCST performance (LS and total errors) significantly differed between groups [F (1,249) = 142, p < .001 and F(1,248) = 26.5, p < .001]. Receiver operating characteristics analysis yielded an AUC of 0.912, sensitivity of 75%, specificity of 73%, and classification accuracy of 84% for NC versus ADCS. Optimal T score cut-offs, regression-based norms, and a T score calculator are provided. Conclusions: These findings support the TCST’s validity and feasibility as a brief measure of executive functioning that can accurately discriminate between normal aging and dementia. Normative data provided will enhance interpretation of this measure.
- 5 February 2026
- Published by Oxford University Press (OUP) in Archives of Clinical Neuropsychology
- Vol. 41 (2)
Texas Card Sorting Test: a Brief Test of Executive Functioning with Age-Adjusted Norms and External Validation
Objective: The Texas Card Sorting Test (TCST) is a brief test of executive functioning that involves sorting cards into groups based on like dimensions. The TCST is shorter than most available executive function measures, requires minimal language demands, and excludes negative feedback. This study generated age-adjusted normative data for the TCST in a healthy aging, normal control (NC) sample, examined its sensitivity and specificity, assessed convergent validity, and compared NC performance to an Alzheimer’s disease clinical syndrome (ADCS) sample. Method: TCST data from 164 NC [Mage = 69.60 (SD = 8.6); Meducation = 15.46 years (SD = 2.83)] and 92 ADCS participants [Mage = 77.73 (SD = 7.55); Meducation = 13.65 years (SD = 3.53)] were obtained as part of a larger neuropsychological battery at Texas Alzheimer’s Research and Care Consortium (TARCC) sites from which age-adjusted normative data were derived and sensitivity, specificity, diagnostic accuracy, and validity were examined. Results: TCST Logical Sorts (LS), or correct card sorts, were significantly correlated with age, education, and other executive function measures. Lower but significant correlations were seen with attention and memory measures. TCST performance (LS and total errors) significantly differed between groups [F (1,249) = 142, p < .001 and F(1,248) = 26.5, p < .001]. Receiver operating characteristics analysis yielded an AUC of 0.912, sensitivity of 75%, specificity of 73%, and classification accuracy of 84% for NC versus ADCS. Optimal T score cut-offs, regression-based norms, and a T score calculator are provided. Conclusions: These findings support the TCST’s validity and feasibility as a brief measure of executive functioning that can accurately discriminate between normal aging and dementia. Normative data provided will enhance interpretation of this measure.
- 8 December 2025
- Published by Oxford University Press (OUP) in Archives of Clinical Neuropsychology
- Vol. 41 (3)
Comparing the Neuropsychology Knowledge Base of Publicly Available Large Language Models
Objective: Assess the neuropsychology knowledge base of open-access large language models (LLMs) and inform potential applications in the field. Method: We obtained 600 multiple-choice practice questions from the “Be Ready for ABPP in Neuropsychology” website of the American Academy of Clinical Neuropsychology. We tested OpenAI (GPT-3.5, GPT-4, o3-mini-high) and Google (Gemini 1.0, 2.0 Flash Thinking Experimental [FTE]) models in two trials (T1: used 20-question blocks, T2: single-item re-administration of incorrect questions during T1). We compared AI-estimated to actual accuracy using a paired-samples t-test, while binomial logit generalized linear mixed-effects models (GLMMs) with a random intercept for item compared LLMs and item-level predictors (domain of practice, word count, position-in-block, and higher/lower order question type). Finally, we thematically analyzed the questions missed by the top models. Results: OpenAI o3-mini-high demonstrated the highest accuracy (T1:87.0%, T2:90.3%), followed by Gemini 2.0 FTE (T1:81.7%, T2:88.7%), GPT-4 (T1:74.0%, T2:85.5%), GPT-3.5 (T1:62.5%), and Gemini 1.0 (T1:52.3%). On average, LLMs overestimated their accuracy by 15.8% (87.3% vs. 71.5%, p < .001). In the GLMMs, specific LLM (p < .001) and practice domain (p = .045) were the only significant predictors of accuracy. Chain-of-thought “reasoning” models outperformed older models (p < .001) but displayed inaccuracies pertaining to aspects of neuropsychological testing and interpretation, diagnostic reasoning, clinical decision making, and neuroanatomy/neuroimaging. Conclusions: Chain-of-thought “reasoning” models displayed the highest accuracy, suggesting they may have utility in neuropsychology education, research, and clinical practice. However, the LLMs displayed persistent neuropsychology content area weaknesses and tended to present inaccurate information with confidence, highlighting the need for caution when interpreting LLM output.
- 8 December 2025
- Published by Oxford University Press (OUP) in Archives of Clinical Neuropsychology
- Vol. 41 (3)
Comparing the Neuropsychology Knowledge Base of Publicly Available Large Language Models
Objective: Assess the neuropsychology knowledge base of open-access large language models (LLMs) and inform potential applications in the field. Method: We obtained 600 multiple-choice practice questions from the “Be Ready for ABPP in Neuropsychology” website of the American Academy of Clinical Neuropsychology. We tested OpenAI (GPT-3.5, GPT-4, o3-mini-high) and Google (Gemini 1.0, 2.0 Flash Thinking Experimental [FTE]) models in two trials (T1: used 20-question blocks, T2: single-item re-administration of incorrect questions during T1). We compared AI-estimated to actual accuracy using a paired-samples t-test, while binomial logit generalized linear mixed-effects models (GLMMs) with a random intercept for item compared LLMs and item-level predictors (domain of practice, word count, position-in-block, and higher/lower order question type). Finally, we thematically analyzed the questions missed by the top models. Results: OpenAI o3-mini-high demonstrated the highest accuracy (T1:87.0%, T2:90.3%), followed by Gemini 2.0 FTE (T1:81.7%, T2:88.7%), GPT-4 (T1:74.0%, T2:85.5%), GPT-3.5 (T1:62.5%), and Gemini 1.0 (T1:52.3%). On average, LLMs overestimated their accuracy by 15.8% (87.3% vs. 71.5%, p < .001). In the GLMMs, specific LLM (p < .001) and practice domain (p = .045) were the only significant predictors of accuracy. Chain-of-thought “reasoning” models outperformed older models (p < .001) but displayed inaccuracies pertaining to aspects of neuropsychological testing and interpretation, diagnostic reasoning, clinical decision making, and neuroanatomy/neuroimaging. Conclusions: Chain-of-thought “reasoning” models displayed the highest accuracy, suggesting they may have utility in neuropsychology education, research, and clinical practice. However, the LLMs displayed persistent neuropsychology content area weaknesses and tended to present inaccurate information with confidence, highlighting the need for caution when interpreting LLM output.
- 23 August 2025
- Published by Taylor & Francis in Applied Neuropsychology: Adult
- Vol. 33 (6) , 2003-2011
Prognostic utility of the Montreal Cognitive Assessment
The Montreal Cognitive Assessment (MoCA) is a widely applied cognitive screening instrument, with a supplemental Memory Index Score (MIS) which has been suggested to predict conversion from Mild Cognitive Impairment (MCI) to Alzheimer's Clinical Syndrome (ACS). This study compared the prognostic utility of the MIS to other MoCA metrics in predicting conversion to ACS or other dementias (OD). We analyzed National Alzheimer's Coordinating Center data from 2900 participants aged 50 years or older, diagnosed with MCI at baseline, with at least one follow-up visit. Multinomial logistic regression models assessed whether baseline MoCA Total Score (TS) or MIS predicted final diagnoses, and receiver operating characteristic (ROC) curves examined the clinical utility of baseline MoCA TS, MIS, Free Recall Score (FRS) and TS+MIS for identifying ACS converters at 1-, 3-, and 5-year follow-ups. Over an average follow-up of 4.65 years, 26.5% converted to ACS and 7.4% to OD. Higher baseline TS was associated with lower odds of conversion to ACS (OR = 0.82) and OD (OR = 0.86), while higher MIS was associated with lower odds of ACS (OR = 0.82) but not OD (OR = 0.97). For identifying ACS, ROC area under the curve ranges showed modest advantage for FRS (0.70-0.73), MIS (0.71-0.74), and TS+MIS (0.70-0.74) over the TS (0.63-0.70). MoCA memory subscores were the strongest baseline indicator of later ACS conversion, but no cut-off score displayed acceptable sensitivity and specificity. Future research may explore if MoCA memory subscores display greater prognostic utility in combination with other ACS features.
- 23 August 2025
- Published by Taylor & Francis in Applied Neuropsychology: Adult
- Vol. 33 (6) , 2003-2011
Prognostic utility of the Montreal Cognitive Assessment
The Montreal Cognitive Assessment (MoCA) is a widely applied cognitive screening instrument, with a supplemental Memory Index Score (MIS) which has been suggested to predict conversion from Mild Cognitive Impairment (MCI) to Alzheimer's Clinical Syndrome (ACS). This study compared the prognostic utility of the MIS to other MoCA metrics in predicting conversion to ACS or other dementias (OD). We analyzed National Alzheimer's Coordinating Center data from 2900 participants aged 50 years or older, diagnosed with MCI at baseline, with at least one follow-up visit. Multinomial logistic regression models assessed whether baseline MoCA Total Score (TS) or MIS predicted final diagnoses, and receiver operating characteristic (ROC) curves examined the clinical utility of baseline MoCA TS, MIS, Free Recall Score (FRS) and TS+MIS for identifying ACS converters at 1-, 3-, and 5-year follow-ups. Over an average follow-up of 4.65 years, 26.5% converted to ACS and 7.4% to OD. Higher baseline TS was associated with lower odds of conversion to ACS (OR = 0.82) and OD (OR = 0.86), while higher MIS was associated with lower odds of ACS (OR = 0.82) but not OD (OR = 0.97). For identifying ACS, ROC area under the curve ranges showed modest advantage for FRS (0.70-0.73), MIS (0.71-0.74), and TS+MIS (0.70-0.74) over the TS (0.63-0.70). MoCA memory subscores were the strongest baseline indicator of later ACS conversion, but no cut-off score displayed acceptable sensitivity and specificity. Future research may explore if MoCA memory subscores display greater prognostic utility in combination with other ACS features.
- Open Access
- 18 August 2025
- Published by SAGE Publications in Journal of Geriatric Psychiatry and Neurology
- Vol. 39 (3) , 261-276
A Scoping Review of Clinical Utility from the Montreal Cognitive Assessment Memory Index Score
Objective: The Montreal Cognitive Assessment (MoCA) Memory Index Score (MIS) is a supplemental assessment of memory composed of word list delayed free-recall followed by step-down category cued- and multiple-choice cued-recall. This paper reviews the MIS literature within Alzheimer’s and other neurodegenerative dementias to synthesize evidence regarding its clinical utility, identify gaps, and inform future research directions. Method: We searched electronic databases of OVID Medline, Embase, PsycINFO, and PubMed from 2014, when the MIS was first described, to July 2025. Peer-reviewed studies that reported data on the diagnostic or prognostic utility of the MIS in assessing neurodegenerative dementia populations were included. Results: We screened 278 articles, and 14 were included in the review. The current literature includes limited reporting on the diagnostic or prognostic utility of the MIS and is characterized by minimal diversity of samples and non-rigorous validation methods. Initial findings are promising and suggestive of incremental validity over the MoCA total score for identifying episodic memory impairment and therefore aiding in differentiation of suspected dementia etiology. However, evidence is insubstantial for the MIS as a tool for predicting progression and additional research is needed to evaluate the incremental validity of the MIS over the conventional MoCA five-word recall score. Conclusions: Large literature gaps exist regarding the clinical utility of the MIS within neurodegenerative dementias. Additional research exploring the psychometric properties of the MIS using diverse samples with rigorous validation methods is needed to better inform its application.
- 18 August 2025
- Published by SAGE Publications in Journal of Geriatric Psychiatry and Neurology
- Vol. 39 (3) , 261-276
Open AccessA Scoping Review of Clinical Utility from the Montreal Cognitive Assessment Memory Index Score
Objective: The Montreal Cognitive Assessment (MoCA) Memory Index Score (MIS) is a supplemental assessment of memory composed of word list delayed free-recall followed by step-down category cued- and multiple-choice cued-recall. This paper reviews the MIS literature within Alzheimer’s and other neurodegenerative dementias to synthesize evidence regarding its clinical utility, identify gaps, and inform future research directions. Method: We searched electronic databases of OVID Medline, Embase, PsycINFO, and PubMed from 2014, when the MIS was first described, to July 2025. Peer-reviewed studies that reported data on the diagnostic or prognostic utility of the MIS in assessing neurodegenerative dementia populations were included. Results: We screened 278 articles, and 14 were included in the review. The current literature includes limited reporting on the diagnostic or prognostic utility of the MIS and is characterized by minimal diversity of samples and non-rigorous validation methods. Initial findings are promising and suggestive of incremental validity over the MoCA total score for identifying episodic memory impairment and therefore aiding in differentiation of suspected dementia etiology. However, evidence is insubstantial for the MIS as a tool for predicting progression and additional research is needed to evaluate the incremental validity of the MIS over the conventional MoCA five-word recall score. Conclusions: Large literature gaps exist regarding the clinical utility of the MIS within neurodegenerative dementias. Additional research exploring the psychometric properties of the MIS using diverse samples with rigorous validation methods is needed to better inform its application.
- 13 February 2024
- Published by Oxford University Press (OUP) in Archives of Clinical Neuropsychology
- Vol. 39 (6) , 732-746
Development and Validity of Norms for Cognitive Dispersion on the Uniform Data Set 3.0 Neuropsychological Battery
Cognitive dispersion indexes intraindividual variability in performance across a battery of neuropsychological tests. Measures of dispersion show promise as markers of cognitive dyscontrol and everyday functioning difficulties; however, they have limited practical applicability due to a lack of normative data. This study aimed to develop and evaluate normed scores for cognitive dispersion among older adults. We analyzed data from 4,283 cognitively normal participants aged ≥50 years from the Uniform Data Set (UDS) 3.0. We describe methods for calculating intraindividual standard deviation (ISD) and coefficient of variation (CoV), as well as associated unadjusted scaled scores and demographically adjusted z-scores. We also examined the ability of ISD and CoV scores to differentiate between cognitively normal individuals (n = 4,283) and those with cognitive impairment due to Lewy body disease (n = 282). We generated normative tables to map raw ISD and CoV scores onto a normal distribution of scaled scores. Cognitive dispersion indices were associated with age, education, and race/ethnicity but not sex. Regression equations were used to develop a freely accessible Excel calculator for deriving demographically adjusted normed scores for ISD and CoV. All measures of dispersion demonstrated excellent diagnostic utility when evaluated by the area under the curve produced from receiver operating characteristic curves. Results of this study provide evidence for the clinical utility of sample-based and demographically adjusted normative standards for cognitive dispersion on the UDS 3.0. These standards can be used to guide interpretation of intraindividual variability among older adults in clinical and research settings.
- 13 February 2024
- Published by Oxford University Press (OUP) in Archives of Clinical Neuropsychology
- Vol. 39 (6) , 732-746
Development and Validity of Norms for Cognitive Dispersion on the Uniform Data Set 3.0 Neuropsychological Battery
Cognitive dispersion indexes intraindividual variability in performance across a battery of neuropsychological tests. Measures of dispersion show promise as markers of cognitive dyscontrol and everyday functioning difficulties; however, they have limited practical applicability due to a lack of normative data. This study aimed to develop and evaluate normed scores for cognitive dispersion among older adults. We analyzed data from 4,283 cognitively normal participants aged ≥50 years from the Uniform Data Set (UDS) 3.0. We describe methods for calculating intraindividual standard deviation (ISD) and coefficient of variation (CoV), as well as associated unadjusted scaled scores and demographically adjusted z-scores. We also examined the ability of ISD and CoV scores to differentiate between cognitively normal individuals (n = 4,283) and those with cognitive impairment due to Lewy body disease (n = 282). We generated normative tables to map raw ISD and CoV scores onto a normal distribution of scaled scores. Cognitive dispersion indices were associated with age, education, and race/ethnicity but not sex. Regression equations were used to develop a freely accessible Excel calculator for deriving demographically adjusted normed scores for ISD and CoV. All measures of dispersion demonstrated excellent diagnostic utility when evaluated by the area under the curve produced from receiver operating characteristic curves. Results of this study provide evidence for the clinical utility of sample-based and demographically adjusted normative standards for cognitive dispersion on the UDS 3.0. These standards can be used to guide interpretation of intraindividual variability among older adults in clinical and research settings.
- 1 February 2024
- Published by Springer Nature in Journal of Health Service Psychology
- Vol. 50 (1) , 37-46
Technology Assistance in Dementia (Tech-AiD): A Framework for Care in the Digital Age
Recent advances in digital technologies hold promise for supporting aging adults and their care partners as they navigate changes in cognitive and daily functioning associated with Alzheimer’s disease and related dementias (ADRD). Commonly owned digital technologies, like smartphones, include features that could help maintain independence and reduce caregiver burden. However, we lack models for successful integration of technologies into treatment of persons with ADRD. We propose the Technology Assistance in Dementia (Tech-AiD) framework for aiding persons with ADRD and their care partners with using digital technologies to reach individualized goals. We discuss how technology use is impacted by a multitude of factors, including severity of cognitive impairment, technology proficiency, and barriers to adequate and equitable care, all of which are further complicated by health disparities. Further, we explore the potential benefits of technology use among patients with ADRD and their care partners, highlighting pertinent clinical and ethical challenges and drawing from evidence-based strategies to promote practical recommendations.
- 1 February 2024
- Published by Springer Nature in Journal of Health Service Psychology
- Vol. 50 (1) , 37-46
Technology Assistance in Dementia (Tech-AiD): A Framework for Care in the Digital Age
Recent advances in digital technologies hold promise for supporting aging adults and their care partners as they navigate changes in cognitive and daily functioning associated with Alzheimer’s disease and related dementias (ADRD). Commonly owned digital technologies, like smartphones, include features that could help maintain independence and reduce caregiver burden. However, we lack models for successful integration of technologies into treatment of persons with ADRD. We propose the Technology Assistance in Dementia (Tech-AiD) framework for aiding persons with ADRD and their care partners with using digital technologies to reach individualized goals. We discuss how technology use is impacted by a multitude of factors, including severity of cognitive impairment, technology proficiency, and barriers to adequate and equitable care, all of which are further complicated by health disparities. Further, we explore the potential benefits of technology use among patients with ADRD and their care partners, highlighting pertinent clinical and ethical challenges and drawing from evidence-based strategies to promote practical recommendations.
- Open Access
- 1 November 2023
- Published by Cambridge University Press (CUP) in Journal of the International Neuropsychological Society
- Vol. 29 (s1) , 482-483
78 Preliminary Exploration of a Novel Speech Analysis Algorithm to Detect Cognitive Impairment in a Spanish Population
Objective:Early detection of mild cognitive impairment (MCI) and dementia is crucial for initiation of treatment and access to appropriate care. While comprehensive neuropsychological assessment is often an intrinsic part of the diagnostic process, access to services may be limited and cannot be utilized effectively on a large scale. For these reasons, cognitive screening instruments are used as brief and cost-effective methods to identify individuals who require further evaluation. Novel technologies and automated software systems to screen for cognitive changes in older individuals are evolving as new avenues for early detection. The present study presents preliminary data on a new technology that uses automated linguistic analysis software to screen for MCI and dementia.Participants and Methods:Data were collected from 148 Spanish-speaking individuals recruited in Spain (MAge=74.4, MEducation=12.93, 56.7% females) of whom 78 were diagnosed as cognitively normal [CN; Mmmse = 28.51 (1.39)], 49 as MCI [MMMSE = 25.65 (2.94)], and 21 as all-cause dementia [MMMSE = 22.52 (2.06)]. Participants were recorded performing various verbal tasks [Animal fluency, phonemic (F) fluency, Cookie Theft Description, and CERAD list learning task]. Recordings were processed via text-transcription and sound signal processing techniques to capture neuropsychological variables and audio characteristics. Features from each task were used in the development of an algorithm (for that task) to compute a score between 0 or 1 (healthy to more impairment), and a fifth algorithm was constructed using audio characteristics from all tasks. These five classifiers were combined algorithmically to provide the final algorithm. Receiver Operating Characteristic (ROC) analysis was conducted to determine sensitivity and specificity of predicted algorithm performance [CN vs. impaired (MCI or dementia)] against clinical diagnoses, and additional general linear modeling was used to test whether age, sex, education, and multilingualism significantly predicted logistically transformed weighted algorithm scores.Results:Scores were transformed to logit scores, with significant differences in mean logit scores between all groups (p <.001). Logit-inverse transformation of mean logit scores (possible range 0 -1) resulted in values of 0.06 for CN, 0.90 for MCI, and 0.99 for all-cause dementia groups. ROC curve analyses revealed the algorithm obtained a total area under the curve of 0.92, with an overall accuracy of 86.8%, a sensitivity of 0.92, and specificity of 0.82. Age was identified as a significant predictor (beta = 0.22; p <0.01) of algorithm output, whereas years of education (beta = -0.04; p = 0.64), sex (beta = 0.38; p = 0.02, did not survive correction for type-1 error), and multilingualism (beta = -0.24; p = 0.22) were non-significant.Conclusions:These findings provide initial support for the utility of an automated speech analysis algorithm to detect cognitive impairment quickly and efficiently in a Spanish-speaking population. Although sociodemographic variables were not included in the algorithm, age significantly predicted algorithm output, and should be further explored to determine if age-adjusted formulas would improve algorithm accuracy for younger versus older individuals. Additional research is needed to validate this novel methodology in other languages, as this may represent a promising cross-cultural screening method for MCI and dementia detection.
- 1 November 2023
- Published by Cambridge University Press (CUP) in Journal of the International Neuropsychological Society
- Vol. 29 (s1) , 482-483
Open Access78 Preliminary Exploration of a Novel Speech Analysis Algorithm to Detect Cognitive Impairment in a Spanish Population
Objective:Early detection of mild cognitive impairment (MCI) and dementia is crucial for initiation of treatment and access to appropriate care. While comprehensive neuropsychological assessment is often an intrinsic part of the diagnostic process, access to services may be limited and cannot be utilized effectively on a large scale. For these reasons, cognitive screening instruments are used as brief and cost-effective methods to identify individuals who require further evaluation. Novel technologies and automated software systems to screen for cognitive changes in older individuals are evolving as new avenues for early detection. The present study presents preliminary data on a new technology that uses automated linguistic analysis software to screen for MCI and dementia.Participants and Methods:Data were collected from 148 Spanish-speaking individuals recruited in Spain (MAge=74.4, MEducation=12.93, 56.7% females) of whom 78 were diagnosed as cognitively normal [CN; Mmmse = 28.51 (1.39)], 49 as MCI [MMMSE = 25.65 (2.94)], and 21 as all-cause dementia [MMMSE = 22.52 (2.06)]. Participants were recorded performing various verbal tasks [Animal fluency, phonemic (F) fluency, Cookie Theft Description, and CERAD list learning task]. Recordings were processed via text-transcription and sound signal processing techniques to capture neuropsychological variables and audio characteristics. Features from each task were used in the development of an algorithm (for that task) to compute a score between 0 or 1 (healthy to more impairment), and a fifth algorithm was constructed using audio characteristics from all tasks. These five classifiers were combined algorithmically to provide the final algorithm. Receiver Operating Characteristic (ROC) analysis was conducted to determine sensitivity and specificity of predicted algorithm performance [CN vs. impaired (MCI or dementia)] against clinical diagnoses, and additional general linear modeling was used to test whether age, sex, education, and multilingualism significantly predicted logistically transformed weighted algorithm scores.Results:Scores were transformed to logit scores, with significant differences in mean logit scores between all groups (p <.001). Logit-inverse transformation of mean logit scores (possible range 0 -1) resulted in values of 0.06 for CN, 0.90 for MCI, and 0.99 for all-cause dementia groups. ROC curve analyses revealed the algorithm obtained a total area under the curve of 0.92, with an overall accuracy of 86.8%, a sensitivity of 0.92, and specificity of 0.82. Age was identified as a significant predictor (beta = 0.22; p <0.01) of algorithm output, whereas years of education (beta = -0.04; p = 0.64), sex (beta = 0.38; p = 0.02, did not survive correction for type-1 error), and multilingualism (beta = -0.24; p = 0.22) were non-significant.Conclusions:These findings provide initial support for the utility of an automated speech analysis algorithm to detect cognitive impairment quickly and efficiently in a Spanish-speaking population. Although sociodemographic variables were not included in the algorithm, age significantly predicted algorithm output, and should be further explored to determine if age-adjusted formulas would improve algorithm accuracy for younger versus older individuals. Additional research is needed to validate this novel methodology in other languages, as this may represent a promising cross-cultural screening method for MCI and dementia detection.
- 19 September 2022
- Published by Wiley in Journal of Neuropsychology
- Vol. 17 (1) , 108-124
The impact of conventional versus robust norming on cognitive characterization and clinical classification of MCI and dementia
We examined the impact of conventional versus robust normative approaches on cognitive characterization and clinical classification of MCI versus dementia. The sample included participants from the National Alzheimer's Coordinating Center Uniform Data Set. Separate demographically adjusted z ‐scores for cognitive tests were derived from conventional ( n = 4273) and robust ( n = 602) normative groups. To assess the impact of deriving scores from a conventional versus robust normative group on cognitive characterization, we examined likelihood of having a low score on each neuropsychological test. Next, we created receiver operating characteristic (ROC) curves for the ability of normed scores derived from each normative group to differentiate between MCI ( n = 3570) and dementia ( n = 1564). We examined the impact of choice of normative group on classification accuracy by comparing sensitivity and specificity values and areas under the curves (AUC). Compared with using a conventional normative group, using a robust normative group resulted in a higher likelihood of low cognitive scores for individuals classified with MCI and dementia. Comparison of the classification accuracy for distinguishing MCI from dementia did not suggest a statistically significant advantage for either normative approach ( Z = −0.29, p = .77; AUC = 0.86 for conventional and AUC = 0.86 for robust). In summary, these results indicate that using a robust normative group increases the likelihood of characterizing cognitive performance as low. However, there is not a clear advantage of using a robust over a conventional normative group when differentiating between MCI and dementia.
- 19 September 2022
- Published by Wiley in Journal of Neuropsychology
- Vol. 17 (1) , 108-124
The impact of conventional versus robust norming on cognitive characterization and clinical classification of MCI and dementia
We examined the impact of conventional versus robust normative approaches on cognitive characterization and clinical classification of MCI versus dementia. The sample included participants from the National Alzheimer's Coordinating Center Uniform Data Set. Separate demographically adjusted z ‐scores for cognitive tests were derived from conventional ( n = 4273) and robust ( n = 602) normative groups. To assess the impact of deriving scores from a conventional versus robust normative group on cognitive characterization, we examined likelihood of having a low score on each neuropsychological test. Next, we created receiver operating characteristic (ROC) curves for the ability of normed scores derived from each normative group to differentiate between MCI ( n = 3570) and dementia ( n = 1564). We examined the impact of choice of normative group on classification accuracy by comparing sensitivity and specificity values and areas under the curves (AUC). Compared with using a conventional normative group, using a robust normative group resulted in a higher likelihood of low cognitive scores for individuals classified with MCI and dementia. Comparison of the classification accuracy for distinguishing MCI from dementia did not suggest a statistically significant advantage for either normative approach ( Z = −0.29, p = .77; AUC = 0.86 for conventional and AUC = 0.86 for robust). In summary, these results indicate that using a robust normative group increases the likelihood of characterizing cognitive performance as low. However, there is not a clear advantage of using a robust over a conventional normative group when differentiating between MCI and dementia.
- Open Access
- 17 August 2022
- Published by Oxford University Press (OUP) in Archives of Clinical Neuropsychology
- Vol. 37 (6) , 1368
A-212 The Limitations of Hippocampal Volumetric Measurements on Clinical Decision-Making: A Case Study
Objective: Integrating multiple pieces of information is a well-known clinical challenge, especially in light of emerging technologies such as NeuroQuant, a fully-automated software used alongside magnetic resonance imaging (MRI) to quantify the volume of brain structures. Using population-based normative data, NeuroQuant compares measurements of certain neuroanatomical features to normal controls, providing metrics for structural abnormalities that may influence the concern for a suspected neurodegenerative etiology. Clinicians should be aware of potential confounding factors when incorporating this information into their clinical impression, as illustrated by this case. Method: We present a unique case study of a 71-year-old Black female who was followed with serial neurological, neuropsychological, and MRI with NeuroQuant evaluations over 10+ years. Results: The patient’s early neuropsychological testing results, in combination with NeuroQuant volumetric measurements, raised suspicion for a potential neurodegenerative process and thereby influenced treatment decisions. However, on longitudinal follow-up, all clinical evaluations were stable except NeuroQuant measurements. From ages 69 to 71, the patient’s hippocampal volumes respectively shifted from the < 5th to 61st normative percentile (see Table 1). Conclusion: Novel volumetric analyses should be interpreted with caution when formulating a clinical impression. Potential explanations for the observed change in NeuroQuant percentile include 1) an update to the normative database, 2) a software update, and 3) the patient moving normative age bands. As with neuropsychological testing, understanding the basis of scores produced by NeuroQuant and characteristics of the normative population is important, as this technology is increasingly used to direct clinical decision-making.
- 17 August 2022
- Published by Oxford University Press (OUP) in Archives of Clinical Neuropsychology
- Vol. 37 (6) , 1368
Open AccessA-212 The Limitations of Hippocampal Volumetric Measurements on Clinical Decision-Making: A Case Study
Objective: Integrating multiple pieces of information is a well-known clinical challenge, especially in light of emerging technologies such as NeuroQuant, a fully-automated software used alongside magnetic resonance imaging (MRI) to quantify the volume of brain structures. Using population-based normative data, NeuroQuant compares measurements of certain neuroanatomical features to normal controls, providing metrics for structural abnormalities that may influence the concern for a suspected neurodegenerative etiology. Clinicians should be aware of potential confounding factors when incorporating this information into their clinical impression, as illustrated by this case. Method: We present a unique case study of a 71-year-old Black female who was followed with serial neurological, neuropsychological, and MRI with NeuroQuant evaluations over 10+ years. Results: The patient’s early neuropsychological testing results, in combination with NeuroQuant volumetric measurements, raised suspicion for a potential neurodegenerative process and thereby influenced treatment decisions. However, on longitudinal follow-up, all clinical evaluations were stable except NeuroQuant measurements. From ages 69 to 71, the patient’s hippocampal volumes respectively shifted from the < 5th to 61st normative percentile (see Table 1). Conclusion: Novel volumetric analyses should be interpreted with caution when formulating a clinical impression. Potential explanations for the observed change in NeuroQuant percentile include 1) an update to the normative database, 2) a software update, and 3) the patient moving normative age bands. As with neuropsychological testing, understanding the basis of scores produced by NeuroQuant and characteristics of the normative population is important, as this technology is increasingly used to direct clinical decision-making.