Machine-Learning Algorithms Based on Screening Tests for Mild Cognitive Impairment

Community screening for dementia among older adults in China: a machine learning-based strategy

BACKGROUND

Dementia is a leading cause of disability in people older than 65 years worldwide. However, diagnosing dementia in its earliest symptomatic stages remains challenging. This study combined specific questions from the AD8 scale with comprehensive health-related characteristics, and used machine learning (ML) to construct diagnostic models of cognitive impairment (CI).

METHODS

The study was based on the Shenzhen Healthy Ageing Research (SHARE) project, and we recruited 823 participants aged 65 years and older, who completed a comprehensive health assessment and cognitive function assessments. Permutation importance was used to select features. Five ML models using BalanceCascade were applied to predict CI: a support vector machine (SVM), multilayer perceptron (MLP), AdaBoost, gradient boosting decision tree (GBDT), and logistic regression (LR). An AD8 score ≥ 2 was used to define CI as a baseline. SHapley Additive exPlanations (SHAP) values were used to interpret the results of ML models.

RESULTS

The first and sixth items of AD8, platelets, waist circumference, body mass index, carcinoembryonic antigens, age, serum uric acid, white blood cells, abnormal electrocardiogram, heart rate, and sex were selected as predictive features. Compared to the baseline (AUC = 0.65), the MLP showed the highest performance (AUC: 0.83 ± 0.04), followed by AdaBoost (AUC: 0.80 ± 0.04), SVM (AUC: 0.78 ± 0.04), GBDT (0.76 ± 0.04). Furthermore, the accuracy, sensitivity and specificity of four ML models were higher than the baseline. SHAP summary plots based on MLP showed the most influential feature on model decision for positive CI prediction was female sex, followed by older age and lower waist circumference.

CONCLUSIONS

The diagnostic models of CI applying ML, especially the MLP, were substantially more effective than the traditional AD8 scale with a score of ≥ 2 points. Our findings may provide new ideas for community dementia screening and to promote such screening while minimizing medical and health resources.

Classification of Mild Cognitive Impairment Using Functional Near-Infrared Spectroscopy-Derived Biomarkers With Convolutional Neural Networks

OBJECTIVE

To date, early detection of mild cognitive impairment (MCI) has mainly depended on paper-based neuropsychological assessments. Recently, biomarkers for MCI detection have gained a lot of attention because of the low sensitivity of neuropsychological assessments. This study proposed the functional near-infrared spectroscopy (fNIRS)-derived data with convolutional neural networks (CNNs) to identify MCI.

METHODS

Eighty-two subjects with MCI and 148 healthy controls (HC) performed the 2-back task, and their oxygenated hemoglobin (HbO2) changes in the prefrontal cortex (PFC) were recorded during the task. The CNN model based on fNIRS-derived spatial features with HbO2 slope within time windows was trained to classify MCI. Thereafter, the 5-fold cross-validation approach was used to evaluate the performance of the CNN model.

RESULTS

Significant differences in averaged HbO2 values between MCI and HC groups were found, and the CNN model could better discriminate MCI with over 89.57% accuracy than the Korean version of the Montreal Cognitive Assessment (MoCA) (89.57%). Specifically, the CNN model based on HbO2 slope within the time window of 20-60 seconds from the left PFC (96.09%) achieved the highest accuracy.

CONCLUSION

These findings suggest that the fNIRS-derived spatial features with CNNs could be a promising way for early detection of MCI as a surrogate for a conventional screening tool and demonstrate the superiority of the fNIRS-derived spatial features with CNNs to the MoCA.

Applications of Artificial Intelligence in the Neuropsychological Assessment of Dementia: A Systematic Review

In the context of advancing healthcare, the diagnosis and treatment of cognitive disorders, particularly Mild Cognitive Impairment (MCI) and Alzheimer's Disease (AD), pose significant challenges. This review explores Artificial Intelligence (AI) and Machine Learning (ML) in neuropsychological assessment for the early detection and personalized treatment of MCI and AD. The review includes 37 articles that demonstrate that AI could be an useful instrument for optimizing diagnostic procedures, predicting cognitive decline, and outperforming traditional tests. Three main categories of applications are identified: (1) combining neuropsychological assessment with clinical data, (2) optimizing existing test batteries using ML techniques, and (3) employing virtual reality and games to overcome the limitations of traditional tests. Despite advancements, the review highlights a gap in developing tools that simplify the clinician's workflow and underscores the need for explainable AI in healthcare decision making. Future studies should bridge the gap between technical performance measures and practical clinical utility to yield accurate results and facilitate clinicians' roles. The successful integration of AI/ML in predicting dementia onset could reduce global healthcare costs and benefit aging societies.

Mental workload classification using convolutional neural networks based on fNIRS-derived prefrontal activity

BACKGROUND

Functional near-infrared spectroscopy (fNIRS) is a tool to assess brain activity during cognitive testing. Despite its usefulness, its feasibility in assessing mental workload remains unclear. This study was to investigate the potential use of convolutional neural networks (CNNs) based on functional near-infrared spectroscopy (fNIRS)-derived signals to classify mental workload in individuals with mild cognitive impairment.

METHODS

Spatial images by constructing a statistical activation map from the prefrontal activity of 120 subjects with MCI performing three difficulty levels of the N-back task (0, 1, and 2-back) were used for CNNs. The CNNs were evaluated using a 5 and 10-fold cross-validation method.

RESULTS

As the difficulty level of the N-back task increased, the accuracy decreased and prefrontal activity increased. In addition, there was a significant difference in the accuracy and prefrontal activity across the three levels (p's < 0.05). The accuracy of the CNNs based on fNIRS-derived spatial images evaluated by 5 and 10-fold cross-validation in classifying the difficulty levels ranged from 0.83 to 0.96.

CONCLUSION

fNIRS could also be a promising tool for measuring mental workload in older adults with MCI despite their cognitive decline. In addition, this study demonstrated the feasibility of the classification performance of the CNNs based on fNIRS-derived signals from the prefrontal cortex.

Sociodemographic Factors Predict Incident Mild Cognitive Impairment: A Brief Review and Empirical Study

OBJECTIVES

Mild cognitive impairment (MCI) is a transitional stage between normal cognitive aging and dementia that increases the risk of progressive cognitive decline. Early prediction of MCI could be beneficial for identifying vulnerable individuals in the community and planning primary and secondary prevention to reduce the incidence of MCI.

DESIGN

A narrative review and cohort study.

SETTING AND PARTICIPANTS

We review the MCI prediction based on the assessment of sociodemographic factors. We included participants from 3 surveys: 8915 from wave 2011/2012 of the China Health and Retirement Longitudinal Study (CHARLS), 9765 from the 2011 Chinese Longitudinal Healthy Longevity Survey (CLHLS), and 1823 from the 2014 Rugao Longevity and Ageing Study (RuLAS).

METHODS

We searched in PubMed, Embase, and Web of Science Core Collection between January 1, 2019, and December 30, 2022. To construct the composite risk score, a multivariate Cox proportional hazards regression model was used. The performance of the score was assessed using receiver operating characteristic (ROC) curves. Furthermore, the composite risk score was validated in 2 longitudinal cohorts, CLHLS and RuLAS.

RESULTS

We concluded on 20 articles from 892 available. The results suggested that the previous models suffered from several defects, including overreliance on cross-sectional data, low predictive utility, inconvenient measurement, and inapplicability to developing countries. Our empirical work suggested that the area under the curve for a 5-year MCI prediction was 0.861 in CHARLS, 0.797 in CLHLS, and 0.823 in RuLAS. We designed a publicly available online tool for this composite risk score.

CONCLUSIONS AND IMPLICATIONS

Attention to these sociodemographic factors related to the incidence of MCI can be beneficially incorporated into the current work, which will set the stage for better early prediction of MCI before its incidence and for reducing the burden of the disease.

In-hospital fall prediction using machine learning algorithms and the Morse fall scale in patients with acute stroke: a nested case-control study

BACKGROUND

Falls are one of the most common accidents in medical institutions, which can threaten the safety of inpatients and negatively affect their prognosis. Herein, we developed a machine learning (ML) model for fall prediction in patients with acute stroke and compared its accuracy with that of the existing fall risk prediction tool, the Morse Fall Scale (MFS).

METHODS

This is a retrospective nested case-control study. The initial sample size was 8462 admitted to a single cerebrovascular specialty hospital with acute stroke. A total of 156 fall events occurred, and each fall case was randomly matched with six control cases. Six ML algorithms were used, namely, regularized logistic regression, support vector machine, naïve Bayes (NB), k-nearest neighbors, random forest, and extreme-gradient boosting (XGB).

RESULTS

We included 156 in the fall group and 934 in the non-fall group. The mean ages of the fall and non-fall groups were 68.3 (± 12.2) and 65.3 (± 12.9) years old, respectively. The MFS total score was significantly higher in the fall group (54.3 ± 18.3) than in the non-fall group (37.7 ± 14.7). The area under the receiver operating curve (AUROC) of the MFS in predicting falls was 0.76 (0.73-0.79). XGB had the highest AUROC of 0.85 (0.78-0.92), and XGB and NB had the highest F1 score of 0.44.

CONCLUSIONS

The AUROC values of all of ML algorithms were similar to those of the MFS in predicting fall risk in patients with acute stroke, allowing for accurate and efficient fall screening.

Can the fNIRS-derived neural biomarker better discriminate mild cognitive impairment than a neuropsychological screening test?

Introduction

Early detection of mild cognitive impairment (MCI), a pre-clinical stage of Alzheimer's disease (AD), has been highlighted as it could be beneficial to prevent progression to AD. Although prior studies on MCI screening have been conducted, the optimized detection way remain unclear yet. Recently, the potential of biomarker for MCI has gained a lot of attention due to a relatively low discriminant power of clinical screening tools.

Methods

This study evaluated biomarkers for screening MCI by performing a verbal digit span task (VDST) using functional near-infrared spectroscopy (fNIRS) to measure signals from the prefrontal cortex (PFC) from a group of 84 healthy controls and 52 subjects with MCI. The concentration changes of oxy-hemoglobin (HbO) were explored during the task in subject groups.

Results

Findings revealed that significant reductions in HbO concentration were observed in the PFC in the MCI group. Specially, the mean of HbO (mHbO) in the left PFC showed the highest discriminant power for MCI, which was higher than that of the Korean version of montreal cognitive assessment (MoCA-K) widely used as a screening tool for MCI. Furthermore, the mHbO in the PFC during the VDST was identified to be significantly correlated to the MoCA-K scores.

Discussion

These findings shed new light on the feasibility and superiority of fNIRS-derived neural biomarker for screening MCI.

Touchscreen Cognitive Tools for Mild Cognitive Impairment and Dementia Used in Primary Care Across Diverse Cultural and Literacy Populations: A Systematic Review

BACKGROUND

Touchscreen cognitive tools opened new promising opportunities for the early detection of cognitive impairment; however, most research studies are conducted in English-speaking populations and high-income countries, with a gap in knowledge about their use in populations with cultural, linguistic, and educational diversity.

OBJECTIVE

To review the touchscreen tools used in primary care settings for the cognitive assessment of mild cognitive impairment (MCI) and dementia, with a focus on populations of different cultures, languages, and literacy.

METHODS

This systematic review was conducted following the PRISMA guidelines. Studies were identified by searching across MEDLINE, EMBASE, EBSCO, OVID, SCOPUS, SCIELO, LILACS, and by cross-referencing. All studies that provide a first-level cognitive assessment for MCI and dementia with any touchscreen tools suitable to be used in the context of primary care were included.

RESULTS

Forty-two studies reporting on 30 tools and batteries were identified. Substantial differences among the tools emerged, in terms of theoretical framework, clinical validity, and features related to the application in clinical practice. A small proportion of the tools are available in multiple languages. Only 7 out of the 30 tools have a multiple languages validation. Only two tools are validated in low-educated samples, e.g., IDEA and mSTS-MCI.

CONCLUSION

General practitioners can benefit from touchscreen cognitive tools. However, easy requirements of the device, low dependence on the examiner, fast administration, and adaptation to different cultures and languages are some of the main features that we need to take into consideration when implementing touchscreen cognitive tools in the culture and language of underrepresented populations.

Can the Virtual Reality-Based Spatial Memory Test Better Discriminate Mild Cognitive Impairment than Neuropsychological Assessment?

Neuropsychological screening tools for mild cognitive impairment (MCI) have been widely used. However, to date, their sensitivity and specificity still remain unsatisfied. This study aims to investigate whether spatial memory can discriminate MCI better than neuropsychological screening tools. A total of 56 healthy older adults and 36 older adults with MCI participated in this study; they performed a spatial cognitive task based on virtual reality (SCT-VR), the Korean version of the Montreal Cognitive Assessment (MoCA-K), and the Wechsler Adult Intelligence Scale-Revised Block Design Test (WAIS-BDT). The discriminant power was compared between the SCT-VR and the MoCA-K, and the reliability and validity of the SCT-VR were analyzed. The spatial memory, assessed by the SCT-VR, showed better sensitivity and specificity (sensitivity: 0.944; specificity: 0.964) than the MoCA-K (sensitivity: 0.857; specificity: 0.746). The test-retest reliability of the SCT-VR was relatively high (ICCs: 0.982, p < 0.001) and the concurrent validity of the SCT-VR with the MoCA-K (r = −0.587, p < 0.001) and the WAIS-BDT (r = −0.594, p < 0.001) was statistically significant. These findings shed light on the importance of spatial memory as a behavioral marker of MCI. The ecologically validated spatial memory tasks based on VR need to be investigated by neuroscientific studies in the future.

Development and validity of computerized neuropsychological assessment devices for screening mild cognitive impairment: Ensemble of models with feature space heterogeneity and retrieval practice effect

OBJECTIVE

This study aimed to develop and validate computerized neuropsychological assessment devices for screening patients with mild cognitive impairment (MCI).

METHODS

We conducted this study in three phases. Phase I involved the development of a conceptual framework of Memory Guard (MG) based on the principles of the cognitive design system (CDS). Phase II involved three steps of feature engineering: item development, filter, and wrapper. Based on the initial items, the number of items in each dimension was determined through analytic hierarchy process. We constructed an initial set with a total of 198 items with three levels of difficulty. Next, we performed feature selection through comprehensive reliability and validity tests, which resulted in the best item bank of 38 test items. The features for modeling were obtained from the best item bank (option scores, reading time scores and total time scores), demographic variables and their MoCA groups. Regarding the heterogeneity of the feature space, we combined the AdaBoost with the Naive Bayes classification algorithm as the decision model of MG. For the screening tool to be used repeatedly, the retrieval practice effect was considered in the design. Phase III involved the validation of measuring instruments. The features incorporated into the modeling process were optimized based on the classification accuracy and area under curve. We also verified the classification effect of the other three classification models with MG.

RESULTS

After three steps of feature engineering, a total of 6 dimensions of cognitive areas were included in MG: orientation, memory, attention, calculation, recall, and language & executive function. 38 features were included in the model (17 features of option score, 20 features of time score, and 1 demographic feature). A total of 333 individuals from two communities in Shanghai and Henan province were included in the measuring instrument verification process. Women accounted for 68.2% of the sample. The median age was 63. 15.3% of the participants had bachelor's degrees or above and 111 participants lived in urban areas (33.3%). The results showed that MG had an accuracy of 93.75% and AUC of 0.923, with a sensitivity of 91.67% and a specificity of 95.45%. Compared to the other three classification models, MG that combined the AdaBoost with the Naive Bayes classification algorithm was the most accurate classifier.

CONCLUSIONS

MG was proved to be reliable and valid in early screening for patients with MCI. MG that integrated heterogeneous features such as demography, option scores, and time scores had a better predictive performance for screening MCI.