Normal Cognition Research Articles

Autism spectrum disorder diagnosis with EEG signals using time series maps of brain functional connectivity and a combined CNN–LSTM model

Background and Objective:People with autism spectrum disorder (ASD) often have cognitive impairments. Effective connectivity between different areas of the brain is essential for normal cognition. Electroencephalography (EEG) has been widely used in the detection of neurological diseases. Previous studies on detecting ASD with EEG data have focused on frequency-related features. Most of these studies have augmented data by splitting the dataset into time slices or sliding windows. However, such approaches to data augmentation may cause the testing data to be contaminated by the training data. To solve this problem, this study developed a novel method for detecting ASD with EEG data. Methods:This study quantified the functional connectivity of the subject’s brain from EEG signals and defined the individual to be the unit of analysis. Publicly available EEG data were gathered from 97 and 92 subjects with ASD and typical development (TD), respectively, while they were at rest or performing a task. Time-series maps of brain functional connectivity were constructed, and the data were augmented using a deep convolutional generative adversarial network. In addition, a combined network for ASD detection, based on convolutional neural network (CNN) and long short-term memory (LSTM), was designed and implemented. Results:Based on functional connectivity, the network achieved classification accuracies of 81.08% and 74.55% on resting state and task state data, respectively. In addition, we found that the functional connectivity of ASD differed from TD primarily in the short-distance functional connectivity of the parietal and occipital lobes and in the distant connections from the right temporoparietal junction region to the left posterior temporal lobe. Conclusions:This paper provides a new perspective for better utilizing EEG to understand ASD. The method proposed in our study is expected to be a reliable tool to assist in the diagnosis of ASD.

Differential atrophy along the longitudinal hippocampal axis in Alzheimer's disease.

Alzheimer's disease (AD) is a progressive neurodegenerative disorder that primarily affects the hippocampus. Since hippocampal studies have highlighted a differential subregional regulation along its longitudinal axis, a more detailed analysis addressing subregional changes along the longitudinal hippocampal axis has the potential to provide new relevant biomarkers. This study included structural brain MRI data of 583 participants from the Alzheimer's Disease Neuroimaging Initiative (ADNI). Cognitively normal (CN) subjects, mild cognitively impaired (MCI) subjects and AD patients were conveniently selected considering the age and sex match between clinical groups. Structural MRI acquisitions were pre-processed and analysed with a new longitudinal axis segmentation method, dividing the hippocampus in three subdivisions (anterior, intermediate, and posterior). When normalizing the volume of hippocampal sub-divisions to total hippocampus, the posterior hippocampus negatively correlates with age only in CN subjects (r= -.31). The longitudinal ratio of hippocampal atrophy (anterior sub-division divided by the posterior one) shows a significant increase with age only in CN (r= .25). Overall, in AD, the posterior hippocampus is predominantly atrophied early on. Consequently, the anterior/posterior hippocampal ratio is an AD differentiating metric at early disease stages with potential for diagnostic and prognostic applications.

Circulating plasma-derived extracellular vesicles expressing bone and kidney markers are associated with neurocognitive impairment in people living with HIV.

Although effective antiretroviral therapy (ART) has improved the life expectancy of people with HIV (PWH), the prevalence of milder forms of HIV-associated neurocognitive disorders (HAND) persist, and it is associated with systemic and neuro-inflammatory processes that could impact other organ systems. However, the complex signaling mechanisms between the bone-kidney systems and the brain in HAND remain unknown. Extracellular vesicles (EVs) play a potential role in inter-organ communication and are involved in regulating cell activity in distant tissues. In this study, we examined whether levels of EVs from bone-and kidney-related cells associate with cognitive dysfunction and explored the relationship between kidney-bone EV axis in PWH experiencing cognitive deficits. EV subtypes were characterized in plasma from 61 PWH with either cognitive impairment (CI, n = 53) or normal cognition (NC, n = 8) based on the American Academy of Neurology criteria for HIV-associated dementia (HAD, n = 11), minor cognitive motor disorder (MCMD, n = 25) or asymptomatic neurocognitive impairment (ANI, n = 17) by spectral flow cytometry. EVs were profiled with markers reflecting bone and kidney cell origin. A support vector machine learning-based model was employed for analyses of EV phenotypes to predict the cognitive dysfunction. Plasma-EVs expressing osteocalcin, sclerostin, and nephrin were significantly higher in the cognitive impairment group compared to the normal cognition group. EVs bearing kidney cell markers correlated significantly with bone-derived EVs. A machine learning-based model, comprised of osteocalcin+, nephrin+, and CD24+ EVs predicted cognitive impairment in PWH on ART. Our study reveals that neurocognitive impairment in PWH is associated with increased levels of plasma EVs enriched with the bone markers osteocalcin and sclerostin and the kidney marker nephrin, suggesting that these EV subtypes may be novel candidate biomarkers for disease-spanning neurocognitive dysfunction. Moreover, the relationship between bone-derived EVs with kidney-derived EVs may suggest their role in mediating inter-organ crosstalk in the pathogenesis of HIV-associated cognitive impairment.

Establishing Reference Values for a New Computerized Cognitive Function Test Program for Children.

To establish reference values for the computerized cognitive test and evaluate cognitive function improvements across different age groups, we introduce the computerized Cognitive Function Test program (eCFT), specifically designed for children. We aimed to establish eCFT reference values and assess cognitive function improvements across different age groups. We included children aged 3-6 years with confirmed normal cognition based on the Korean Developmental Screening Test for Infants and Children and Kaufman Assessment Battery for Children-II. The eCFT consists of 8 subtests for visual perception, attention, memory, and executive function. A total of 66 participants (36 males and 30 females) with an average age of 4.4 years participated. The age 6 group consistently outperformed both age group 3 and 4 in terms of correct responses. With regard to the completed stage, the "selective auditory stimulus" test findings were 2.0 and 3.9 for the age 3 and age 6 groups, respectively (p<0.05). The "trail-making" test findings were 1.7, 2.1, 2.6, and 2.8, respectively (between ages 3 and 6, p<0.01; between ages 4 and 6, p<0.05); moreover, the age 5 group surpassed the age 3 group (2.6 and 1.7, respectively, p<0.05). The eCFT is an easily accessible tool to evaluate cognitive function in young children. We introduce reference values with a cutoff range for preschool-aged children, enabling early intervention for those with cognitive impairment. Given its accessibility and relatively short evaluation time, the eCFT has potential for clinical use.

Pathologic and cognitive correlates of plasma biomarkers in neurodegenerative disease.

We investigate pathological correlates of plasma phosphorylated tau 181 (p-tau181), glial fibrillary acidic protein (GFAP), and neurofilament light chain (NfL) across a clinically diverse spectrum of neurodegenerative disease, including normal cognition (NormCog) and impaired cognition (ImpCog). Participants were NormCog (n=132) and ImpCog (n=461), with confirmed β-amyloid (Aβ+/-) status (cerebrospinal fluid, positron emission tomography, autopsy) and single molecule array plasma measurements. Logistic regression and receiver operating characteristic (ROC) area under the curve (AUC) tested how combining plasma analytes discriminated Aβ+ from Aβ-. Survival analyses tested time to clinical dementia rating (global CDR) progression. Multivariable models (p-tau+GFAP+NfL) had the best performance to detect Aβ+ in NormCog (ROCAUC=0.87) and ImpCog (ROCAUC=0.87). Survival analyses demonstrated that higher NfL best predicted faster CDR progression for both Aβ+ (hazard ratio [HR]=2.94; p=8.1e-06) and Aβ- individuals (HR=3.11; p=2.6e-09). Combining plasma biomarkers can optimize detection of Alzheimer's disease (AD) pathology across cognitively normal and clinically diverse neurodegenerative disease. Participants were clinically heterogeneous, with autopsy- or biomarker-confirmed Aβ. Combining plasma p-tau181, GFAP, and NfL improved diagnostic accuracy for Aβ status. Diagnosis by plasma biomarkers is more accurate in amnestic AD than nonamnestic AD. Plasma analytes show independent associations with tau PET and post mortem Aβ/tau. Plasma NfL predicted longitudinal cognitive decline in both Aβ+ and Aβ- individuals.

Data-driven clustering approach to identify novel clusters of high cognitive impairment risk among Chinese community-dwelling elderly people with normal cognition: A national cohort study.

Cognitive impairment is a highly heterogeneous disorder that necessitates further investigation into the distinct characteristics of populations at varying risk levels of cognitive impairment. Using a large-scale registry cohort of elderly individuals, we applied a data-driven approach to identify novel clusters based on diverse sociodemographic features. A prospective cohort of 6398 elderly people from the Chinese Longitudinal Healthy Longevity Survey, followed between 2008-14, was used to develop and validate the model. Participants were aged ≥60 years, community-dwelling, and the Chinese version of the Mini-Mental State Examination (MMSE) score ≥18 were included. Sixty-nine sociodemographic features were included in the analysis. The total population was divided into two-thirds for the derivation cohort (n = 4265) and one-third for the validation cohort (n = 2133). In the derivation cohort, an unsupervised Gaussian mixture model was applied to categorise participants into distinct clusters. A classifier was developed based on the most important 10 factors and was applied to categorise participants into their corresponding clusters in a validation cohort. The difference in the three-year risk of cognitive impairment was compared across the clusters. We identified four clusters with distinct features in the derivation cohort. Cluster 1 was associated with the worst life independence, longest sleep duration, and the oldest age. Cluster 2 demonstrated the highest loneliness, characterised by non-marital status and living alone. Cluster 3 was characterised by the lowest sense of loneliness and the highest proportions in marital status and family co-residence. Cluster 4 demonstrated heightened engagement in exercise and leisure activity, along with independent decision-making, hygiene, and a diverse diet. In comparison to Cluster 4, Cluster 1 exhibited the highest three-year cognitive impairment risk (adjusted odds ratio (aOR) = 3.31; 95% confidence interval (CI) = 1.81-6.05), followed by Cluster 2 and Cluster 3 after adjustment for baseline MMSE, residence, sex, age, years of education, drinking, smoking, hypertension, diabetes, heart disease and stroke or cardiovascular diseases. A data-driven approach can be instrumental in identifying individuals at high risk of cognitive impairment among cognitively normal elderly populations. Based on various sociodemographic features, these clusters can suggest individualised intervention plans.

Integrating Demographics and Imaging Features for Various Stages of Dementia Classification: Feed Forward Neural Network Multi-Class Approach.

MRI magnetization-prepared rapid acquisition (MPRAGE) is an easily available imaging modality for dementia diagnosis. Previous studies suggested that volumetric analysis plays a crucial role in various stages of dementia classification. In this study, volumetry, radiomics and demographics were integrated as inputs to develop an artificial intelligence model for various stages, including Alzheimer's disease (AD), mild cognitive decline (MCI) and cognitive normal (CN) dementia classifications. The Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset was separated into training and testing groups, and the Open Access Series of Imaging Studies (OASIS) dataset was used as the second testing group. The MRI MPRAGE image was reoriented via statistical parametric mapping (SPM12). Freesurfer was employed for brain segmentation, and 45 regional brain volumes were retrieved. The 3D Slicer software was employed for 107 radiomics feature extractions from within the whole brain. Data on patient demographics were collected from the datasets. The feed-forward neural network (FFNN) and the other most common artificial intelligence algorithms, including support vector machine (SVM), ensemble classifier (EC) and decision tree (DT), were used to build the models using various features. The integration of brain regional volumes, radiomics and patient demographics attained the highest overall accuracy at 76.57% and 73.14% in ADNI and OASIS testing, respectively. The subclass accuracies in MCI, AD and CN were 78.29%, 89.71% and 85.14%, respectively, in ADNI testing, as well as 74.86%, 88% and 83.43% in OASIS testing. Balanced sensitivity and specificity were obtained for all subclass classifications in MCI, AD and CN. The FFNN yielded good overall accuracy for MCI, AD and CN categorization, with balanced subclass accuracy, sensitivity and specificity. The proposed FFNN model is simple, and it may support the triage of patients for further confirmation of the diagnosis.

Performance Assessment of ChatGPT versus Bard in Detecting Alzheimer's Dementia.

Large language models (LLMs) find increasing applications in many fields. Here, three LLM chatbots (ChatGPT-3.5, ChatGPT-4, and Bard) are assessed in their current form, as publicly available, for their ability to recognize Alzheimer's dementia (AD) and Cognitively Normal (CN) individuals using textual input derived from spontaneous speech recordings. A zero-shot learning approach is used at two levels of independent queries, with the second query (chain-of-thought prompting) eliciting more detailed information than the first. Each LLM chatbot's performance is evaluated on the prediction generated in terms of accuracy, sensitivity, specificity, precision, and F1 score. LLM chatbots generated a three-class outcome ("AD", "CN", or "Unsure"). When positively identifying AD, Bard produced the highest true-positives (89% recall) and highest F1 score (71%), but tended to misidentify CN as AD, with high confidence (low "Unsure" rates); for positively identifying CN, GPT-4 resulted in the highest true-negatives at 56% and highest F1 score (62%), adopting a diplomatic stance (moderate "Unsure" rates). Overall, the three LLM chatbots can identify AD vs. CN, surpassing chance-levels, but do not currently satisfy the requirements for clinical application.

News event memory in amnestic and non-amnestic MCI, heritable risk for dementia, and subjective memory complaints

Robust and sensitive clinical measures are needed for more accurate and earlier detection of Alzheimer's disease (AD), for staging preclinical AD, and for gauging the efficacy of treatments. Mild impairment on episodic memory tests is thought to indicate a cognitive risk of developing AD and mild cognitive impairment (MCI), considered to be a transitional stage between normal aging and AD. Novel tests of semantic memory, such as memory for news events, are also impaired early on but have received little clinical attention even though they may provide a novel way to assess cognitive risk for AD. We examined memory for news events in older adults with normal cognition (NC, N = 34), amnestic MCI (aMCI, N = 27), or non-aMCI (N = 10) using the Retrograde Memory News Events Test (RM-NET). We asked if news event memory was sensitive to 1) aMCI and also non-aMCI, which has rarely been examined, 2) genetic risk for dementia (positive family history of any type of dementia, presence of an APOE-4 allele, or polygenic risk for AD), and 3) subjective memory functioning judgments about the past. We found that both MCI subgroups exhibited impaired RM-NET Lifespan accuracy scores together with temporally-limited retrograde amnesia. For the aMCI group amnesia extended back 45 years prior to testing, but not beyond that time frame. The extent of retrograde amnesia could not be reliably estimated in the small non-aMCI group. The effect sizes of having MCI on the RM-NET were medium for the non-aMCI group and large for the aMCI group, whereas the effect sizes of participant characteristics on RM-NET accuracy scores were small. For the combined MCI group (N = 37), news event memory was significantly related to positive family history of dementia but was not related to the more specific genetic markers of AD risk. For the NC group, news event memory was not related to any measure of genetic risk. Objective measures of past memory from the RM-NET were not related to subjective memory judgements about the present or the recent past in either group. By contrast, when individuals subjectively compared their present versus past memory abilities, there was a significant association between this judgment and objective measures of the past from the RM-NET (direct association for the NC group and inverse for the MCI group). The RM-NET holds significant promise for early identification of those with cognitive and genetic risk factors for AD and non-AD dementias.

Prediction of Cognitive Progression Due to Alzheimer's Disease in Normal Participants Based on Individual Default Mode Network Metabolic Connectivity Strength

BackgroundPredicting cognitive decline among individuals in the aging population who are already amyloid-β (Aβ) positive or tau positive poses clinical challenges. In Alzheimer's disease research, intra–default mode network (DMN) connections play a pivotal role in diagnosis. In this article, we propose metabolic connectivity within the DMN as a supplementary biomarker to the Aβ, pathological tau, and neurodegeneration framework. MethodsExtracting data from 1292 participants in the Alzheimer’s Disease Neuroimaging Initiative, we collected paired T1-weighted structural magnetic resonance imaging and 18F-labeled-fluorodeoxyglucose positron emission computed tomography scans. Individual metabolic DMN networks were constructed, and metabolic connectivity (MC) strength in the DMN was assessed. In the cognitively unimpaired group, the Cox model identified cognitively unimpaired (MC+), high-risk participants, with Kaplan-Meier survival analyses and hazard ratios revealing the strength of MC’s predictive performance. Spearman correlation analyses explored relationships between MC strength, and Aβ, pathological tau, neurodegeneration biomarkers, and clinical scales. DMN standard uptake value ratio (SUVR) provided comparative insights in the analyses. ResultsBoth MC strength and SUVR exhibited gradual declines with cognitive deterioration, displaying significant intergroup differences. Survival analyses indicated enhanced Aβ and tau prediction with both metrics, with MC strength outperforming SUVR. Combined MC strength and Aβ yielded optimal predictive performance (hazard ratio = 9.29), followed by MC strength and tau (hazard ratio = 8.92). Generally, the strength of MC’s correlations with Aβ, pathological tau, and neurodegeneration biomarkers exceeded SUVR. ConclusionsIndividuals with normal cognition and disrupted DMN metabolic connectivity face an elevated risk of cognitive decline linked to Aβ that precedes metabolic issues.

Comparison between morphometry and radiomics: detecting normal brain aging based on grey matter.

Voxel-based morphometry (VBM), surface-based morphometry (SBM), and radiomics are widely used in the field of neuroimage analysis, while it is still unclear that the performance comparison between traditional morphometry and emerging radiomics methods in diagnosing brain aging. In this study, we aimed to develop a VBM-SBM model and a radiomics model for brain aging based on cognitively normal (CN) individuals and compare their performance to explore both methods' strengths, weaknesses, and relationships. 967 CN participants were included in this study. Subjects were classified into the middle-aged group (n = 302) and the old-aged group (n = 665) according to the age of 66. The data of 360 subjects from the Alzheimer's Disease Neuroimaging Initiative were used for training and internal test of the VBM-SBM and radiomics models, and the data of 607 subjects from the Australian Imaging, Biomarker and Lifestyle, the National Alzheimer's Coordinating Center, and the Parkinson's Progression Markers Initiative databases were used for the external tests. Logistics regression participated in the construction of both models. The area under the receiver operating characteristic curve (AUC), sensitivity, specificity, accuracy, positive predictive value, and negative predictive value were used to evaluate the two model performances. The DeLong test was used to compare the differences in AUCs between models. The Spearman correlation analysis was used to observe the correlations between age, VBM-SBM parameters, and radiomics features. The AUCs of the VBM-SBM model and radiomics model were 0.697 and 0.778 in the training set (p = 0.018), 0.640 and 0.789 in the internal test set (p = 0.007), 0.736 and 0.737 in the AIBL test set (p = 0.972), 0.746 and 0.838 in the NACC test set (p < 0.001), and 0.701 and 0.830 in the PPMI test set (p = 0.036). Weak correlations were observed between VBM-SBM parameters and radiomics features (p < 0.05). The radiomics model achieved better performance than the VBM-SBM model. Radiomics provides a good option for researchers who prioritize performance and generalization, whereas VBM-SBM is more suitable for those who emphasize interpretability and clinical practice.

Can Videoconferencing Be a Viable Method to Measure Speech Perception?

Telehealth has proven effective for service delivery to remote and rural locations and was helpful during lockdowns when patients were unable to see clinicians in person. To assure the reliability of clinical services, the aim of the present study was to evaluate a telehealth protocol to measure speech perception skills through virtual meeting applications such as Zoom. A total of 20 participants with normal hearing and cognition participated in the study. Participants' speech perception performance was measured in two sessions: one over a Zoom call and one in person in the laboratory. Speech perception was measured using Quick Speech-in-Noise Test, Bamford-Kowal-Bench Speech-in-Noise Test, and Words-in-Noise Test. Statistical analysis revealed no significant differences between speech perception scores obtained over the Zoom and in-person methodologies for any of the tests. Additionally, our study found no significant difference in the scores obtained between wired and wireless headphones during Zoom calls. These results suggest that Zoom can be used as a reliable method to measure speech perception in young individuals with normal hearing using these three tests in situations where conventional methods cannot be utilized.

Prefrontal intra-individual ERP variability and its asymmetry: exploring its biomarker potential in mild cognitive impairment

BackgroundThe worldwide trend of demographic aging highlights the progress made in healthcare, albeit with health challenges like Alzheimer’s Disease (AD), prevalent in individuals aged 65 and above. Its early detection at the mild cognitive impairment (MCI) stage is crucial. Event-related potentials (ERPs) obtained by averaging EEG segments responded to repeated events are vital for cognitive impairment research. Consequently, examining intra-trial ERP variability is vital for comprehending fluctuations within psychophysiological processes of interest. This study aimed to investigate cognitive deficiencies and instability in MCI using ERP variability and its asymmetry from a prefrontal two-channel EEG device.MethodsIn this study, ERP variability for both target and non-target responses was examined using the response variance curve (RVC) in a sample comprising 481 participants with MCI and 1,043 age-matched healthy individuals. The participants engaged in auditory selective attention tasks. Cognitive decline was assessed using the Seoul Neuropsychological Screening Battery (SNSB) and the Mini-Mental State Examination (MMSE). The research employed various statistical methods, including independent t-tests, and univariate and multiple logistic regression analyses. These analyses were conducted to investigate group differences and explore the relationships between neuropsychological test results, ERP variability and its asymmetry measures, and the prevalence of MCI.ResultsOur results showed that patients with MCI exhibited unstable cognitive processing, characterized by increased ERP variability compared to cognitively normal (CN) adults. Multiple logistic regression analyses confirmed the association between ERP variability in the target and non-target responses with MCI prevalence, independent of demographic and neuropsychological factors.DiscussionThe unstable cognitive processing in the MCI group compared to the CN individuals implies abnormal neurological changes and reduced and (or) unstable attentional maintenance during cognitive processing. Consequently, utilizing ERP variability measures from a portable EEG device could serve as a valuable addition to the conventional ERP measures of latency and amplitude. This approach holds significant promise for identifying mild cognitive deficits and neural alterations in individuals with MCI.

A multi-omics study to monitor senescence-associated secretory phenotypes of Alzheimer's disease.

Alzheimer's disease (AD) is characterized by the progressive degeneration and damage of neurons in the brain. However, developing an accurate diagnostic assay using blood samples remains a challenge in clinic practice. The aim of this study was to explore senescence-associated secretory phenotypes (SASPs) in peripheral blood using mass spectrometry based multi-omics approach and to establish diagnostic assays for AD. This retrospective study included 88 participants, consisting of 29 AD patients and 59 cognitively normal (CN) individuals. Plasma and serum samples were examined using high-resolution mass spectrometry to identify proteomic and metabolomic profiles. Receiver operating characteristic (ROC) analysis was employed to screen biomarkers with diagnostic potential. K-nearest neighbors (KNN) algorithm was utilized to construct a multi-dimensional model for distinguishing AD from CN. Proteomics analysis revealed upregulation of five plasma proteins in AD, including RNA helicase aquarius (AQR), zinc finger protein 587B (ZNF587B), C-reactive protein (CRP), fibronectin (FN1), and serum amyloid A-1 protein (SAA1), indicating their potential for AD classification. Interestingly, KNN-based three-dimensional model, comprising AQR, ZNF587B, and CRP, demonstrated its high accuracy in AD recognition, with evaluation possibilities of 0.941, 1.000, and 1.000 for the training, testing, and validation datasets, respectively. Besides, metabolomics analysis suggested elevated levels of serum phenylacetylglutamine (PAGIn) in AD. The multi-omics outcomes highlighted the significance of the SASPs, specifically AQR, ZNF587B, CRP, and PAGIn, in terms of their potential for diagnosing AD and suggested neuronal aging-associated pathophysiology.

Free-Water Imaging of the Nucleus Basalis of Meynert in Patients With Idiopathic REM Sleep Behavior Disorder and Parkinson Disease.

Cognitive impairments are common in idiopathic REM sleep behavior disorder (iRBD), in which the cholinergic degeneration of nucleus basalis of Meynert (NBM) may play an important role. However, the progressive changes of NBM, the relationship between progressive NBM degeneration and progression of cognitive impairments, and whether degeneration of the NBM can predict cognitive decline in patients with iRBD remain unclear. This study aimed to investigate the cross-sectional and longitudinal microstructural alterations in the NBM of patients with iRBD using free-water imaging and whether free water in the NBM is related to cognitive, mood, and autonomic function. We compared the baseline free-water values in the NBM between 59 healthy controls (HCs), 57 patients with iRBD, 57 patients with Parkinson disease (PD) with normal cognition (PD-NC), and 64 patients with PD with cognitive impairment (PD-CI). Thirty patients with iRBD and 40 HCs had one longitudinal data. In patients with iRBD, we explored the associations between baseline and longitudinal changes of free-water values in the NBM and clinical characteristics and whether baseline free-water values in the NBM could predict cognitive decline. IRBD, PD-NC, and PD-CI groups had significantly increased free-water values in the NBM compared with HCs, whereas PD-CI had higher free-water values compared with iRBD and PD-NC. In patients with iRBD, free-water values in the NBM were progressively elevated over follow-up and correlated with the progression of cognitive impairment and depression. Free-water values in the NBM could predict cognitive decline in the iRBD group. Furthermore, we found that patients with iRBD with cognitive impairment had higher relative change of free-water value in the NBM compared with patients with iRBD with normal cognition over follow-up. This study proves that free-water values in the NBM are elevated cross-sectionally and longitudinally and are associated with the progression of cognitive impairment and depression in patients with iRBD. Moreover, the free-water value in the NBM can predict cognitive decline in patients with iRBD. Whether free-water imaging of the NBM has the potential to be a marker for monitoring progressive cognitive impairment and predicting the conversion to dementia in synucleinopathies needs further investigation.

Cholinergic hyperintensity pathways are associated with cognitive performance in patients with asymptomatic carotid artery stenosis

ObjectivesWe aimed to determine whether asymptomatic carotid artery stenosis (ACS) induced cognitive impairments were related to the cholinergic hyperintensity pathway. MethodsThis cross-sectional study included patients with moderate-to-severe ACS, who were categorized into mild cognitive impairment (MCI) and normal cognition groups on the basis of Montreal Cognitive Assessment (MoCA) scores. The cholinergic pathway hyperintensity scale (CHIPS), Fazekas, and medial temporal atrophy (MTA) scores were assessed. SPSS software was used for statistical analyses. ResultsA total of 117 ACS patients (70.89 ± 8.81 years) and 105 controls (67.87 ± 9.49 years) were evaluated (t = 2.46, p = 0.015). The ACS group showed a worse median Mini-Mental Status Examination (MMSE) score (z = −2.41, p = 0.016) and MoCA score (z = −3.51, p < 0.001), and a significantly higher median total CHIPS score (z = 4.88, p < 0.001) and mean Fazekas score (t = 2.39, p = 0.018). In the correlation analysis, the MoCA score showed a significant negative correlation with the CHIPS score (ρ = −0.41, p < 0.001) and Fazekas score (ρ = −0.31, p < 0.001) in ACS group. Logistic regression analyses suggested that CHIPS scores were risk factors for MCI in patients with ACS (odds ratio [OR] = 1.07, 95% Confidence Interval [CI]1.01–1.13 and controls (OR = 1.09, 95%CI 1.01–1.17), while the MTA and Fazekas scores showed no predictive power. The receiver operating characteristic curve showed that the area under the curve of the CHIPS score for predicting MCI was 0.71 in ACS group, but was only 0.57 in controls. ConclusionsPatients with ACS showed poorer cognitive performance and higher CHIPS and Fazekas scores. CHIPS, but not Fazekas, scores were risk factors for cognitive impairment and were a valuable factor to predict MCI in patients with ACS.

Performance Evaluation of Deep, Shallow and Ensemble Machine Learning Methods for the Automated Classification of Alzheimer's Disease.

Artificial intelligence (AI)-based approaches are crucial in computer-aided diagnosis (CAD) for various medical applications. Their ability to quickly and accurately learn from complex data is remarkable. Deep learning (DL) models have shown promising results in accurately classifying Alzheimer's disease (AD) and its related cognitive states, Early Mild Cognitive Impairment (EMCI) and Late Mild Cognitive Impairment (LMCI), along with the healthy conditions known as Cognitively Normal (CN). This offers valuable insights into disease progression and diagnosis. However, certain traditional machine learning (ML) classifiers perform equally well or even better than DL models, requiring less training data. This is particularly valuable in CAD in situations with limited labeled datasets. In this paper, we propose an ensemble classifier based on ML models for magnetic resonance imaging (MRI) data, which achieved an impressive accuracy of 96.52%. This represents a 3-5% improvement over the best individual classifier. We evaluated popular ML classifiers for AD classification under both data-scarce and data-rich conditions using the Alzheimer's Disease Neuroimaging Initiative and Open Access Series of Imaging Studies datasets. By comparing the results to state-of-the-art CNN-centric DL algorithms, we gain insights into the strengths and weaknesses of each approach. This work will help users to select the most suitable algorithm for AD classification based on data availability.

Data-driven classification of cognitively normal and mild cognitive impairment subtypes predicts progression in the NACC dataset.

Data-driven neuropsychological methods can identify mild cognitive impairment (MCI) subtypes with stronger associations to dementia risk factors than conventional diagnostic methods. Cluster analysis used neuropsychological data from participants without dementia (mean age=71.6 years) in the National Alzheimer's Coordinating Center (NACC) Uniform Data Set (n=26,255) and the "normal cognition" subsample (n=16,005). Survival analyses examined MCI or dementia progression. Five clusters were identified: "Optimal" cognitively normal (oCN; 13.2%), "Typical" CN (tCN; 28.0%), Amnestic MCI (aMCI; 25.3%), Mixed MCI-Mild (mMCI-Mild; 20.4%), and Mixed MCI-Severe (mMCI-Severe; 13.0%). Progression to dementia differed across clusters (oCN<tCN<aMCI<mMCI-Mild<mMCI-Severe). Cluster analysis identified more MCI cases than consensus diagnosis. In the "normal cognition" subsample, five clusters emerged: High-All Domains (High-All; 16.7%), Low-Attention/Working Memory (Low-WM; 22.1%), Low-Memory (36.3%), Amnestic MCI (16.7%), and Non-amnestic MCI (naMCI; 8.3%), with differing progression rates (High-All<Low-WM=Low-Memory<aMCI<naMCI). Our data-driven methods outperformed consensus diagnosis by providing more precise information about progression risk and revealing heterogeneity in cognition and progression risk within the NACC "normal cognition" group.

The characteristic patterns of individual brain susceptibility networks underlie Alzheimer’s disease and white matter hyperintensity-related cognitive impairment

Excessive iron accumulation in the brain cortex increases the risk of cognitive deterioration. However, interregional relationships (defined as susceptibility connectivity) of local brain iron have not been explored, which could provide new insights into the underlying mechanisms of cognitive decline. Seventy-six healthy controls (HC), 58 participants with mild cognitive impairment due to probable Alzheimer’s disease (MCI-AD) and 66 participants with white matter hyperintensity (WMH) were included. We proposed a novel approach to construct a brain susceptibility network by using Kullback‒Leibler divergence similarity estimation from quantitative susceptibility mapping and further evaluated its topological organization. Moreover, sparse logistic regression (SLR) was applied to classify MCI-AD from HC and WMH with normal cognition (WMH-NC) from WMH with MCI (WMH-MCI).The altered susceptibility connectivity in the MCI-AD patients indicated that relatively more connectivity was involved in the default mode network (DMN)-related and visual network (VN)-related connectivity, while more altered DMN-related and subcortical network (SN)-related connectivity was found in the WMH-MCI patients. For the HC vs. MCI-AD classification, the features selected by the SLR were primarily distributed throughout the DMN-related and VN-related connectivity (accuracy = 76.12%). For the WMH-NC vs. WMH-MCI classification, the features with high appearance frequency were involved in SN-related and DMN-related connectivity (accuracy = 84.85%). The shared and specific patterns of the susceptibility network identified in both MCI-AD and WMH-MCI may provide a potential diagnostic biomarker for cognitive impairment, which could enhance the understanding of the relationships between brain iron burden and cognitive decline from a network perspective.

The Hippocampal Subfield Volume Reduction and Plasma Biomarker Changes in Mild Cognitive Impairment and Alzheimer's Disease.

The hippocampus consists of histologically and functionally distinct subfields, which shows differential vulnerabilities to Alzheimer's disease (AD)-associated pathological changes. To investigate the atrophy patterns of the main hippocampal subfields in patients with mild cognitive impairment (MCI) and AD and the relationships among the hippocampal subfield volumes, plasma biomarkers and cognitive performance. This cross-sectional study included 119 patients stratified into three categories: normal cognition (CN; N = 40), MCI (N = 39), and AD (N = 40). AD-related plasma biomarkers were measured, including amyloid-β (Aβ)42, Aβ40, Aβ42/Aβ40 ratio, p-tau181, and p-tau217, and the hippocampal subfield volumes were calculated using automated segmentation and volumetric procedures implemented in FreeSurfer. The subiculum body, cornu ammonis (CA) 1-head, CA1-body, CA4-body, molecular_layer_HP-head, molecular_layer_HP-body, and GC-ML-DG-body volumes were smaller in the MCI group than in the CN group. The subiculum body and CA1-body volumes accurately distinguished MCI from CN (area under the curve [AUC] = 0.647-0.657). The subiculum-body, GC-ML-DG-body, CA4-body, and molecular_layer_HP-body volumes accurately distinguished AD from MCI (AUC = 0.822-0.833) and AD from CN (AUC = 0.903-0.905). The p-tau 217 level served as the best plasma indicator of AD and correlated with broader hippocampal subfield volumes. Moreover, mediation analysis demonstrated that the subiculum-body volume mediated the associations between the p-tau217 and p-tau181 levels, and the Montreal Cognitive Assessment and Auditory Verbal Learning Test recognition scores. Hippocampal subfields with distinctive atrophy patterns may mediate the effects of tau pathology on cognitive function. The subiculum-body may be the most clinically meaningful hippocampal subfield, which could be an effective target region for assessing disease progression.