Search

Background Individualized and reliable biomarkers are crucial for diagnosing Alzheimer’s disease (AD). However, lack of accessibility and neurobiological correlation are the main obstacles to their clinical application. Machine learning algorithms can effectively identify personalized biomarkers based on the prominent symptoms of AD. Methods Episodic memory–related magnetic resonance imaging (MRI) features of 143 patients with amnesic mild cognitive impairment (MCI) were identified using a multivariate relevance vector regression algorithm. The support vector machine classification model was constructed using these MRI features and verified in 2 independent datasets (N = 994). The neurobiological basis was also investigated based on cognitive assessments, neuropathologic biomarkers of cerebrospinal fluid, and positron emission tomography images of amyloid-β plaques. Results The combination of gray matter volume and amplitude of low-frequency fluctuation MRI features accurately predicted episodic memory impairment in individual patients with amnesic MCI (r = .638) when measured using an episodic memory assessment panel. The MRI features that contributed to episodic memory prediction were primarily distributed across the default mode network and limbic network. The classification model based on these features distinguished patients with AD from normal control subjects with more than 86% accuracy. Furthermore, most identified episodic memory–related regions showed significantly different amyloid-β positron emission tomography measurements among the AD, MCI, and normal control groups. Moreover, the classification outputs significantly correlated with cognitive assessment scores and cerebrospinal fluid pathological biomarkers' levels in the MCI and AD groups. Conclusions Neuroimaging features can reflect individual episodic memory function and serve as potential diagnostic biomarkers of AD.

We examined the characteristics of individuals with biomarker evidence of tauopathy but without β-amyloid (Aβ) (A-T+) in relation to individuals with (A+T+) and without (A-T-) evidence of Alzheimer’s disease (AD). We included 561 participants with Aβ and tau PET from the Alzheimer’s Disease Neuroimaging Initiative (ADNI). We compared A-T- (n = 316), A-T+ (n = 63), and A+T+ (n = 182) individuals on demographics, amyloid, tau, hippocampal volumes, and cognition. A-T+ individuals were low on apolipoprotein E ɛ4 prevalence (17%) and had no evidence of subtly elevated brain Aβ within the negative range. The severity of tau deposition, hippocampal atrophy, and cognitive dysfunction in the A-T+ group was intermediate between A-T- and A+T+ (all p < 0.001). Tau uptake patterns in A-T+ individuals were heterogeneous, but approximately 29% showed tau deposition in the medial temporal lobe only, consistent with primary age-related tauopathy and an additional 32% showed a pattern consistent with AD. A-T+ individuals also share other features that are characteristic of AD such as cognitive impairment and neurodegeneration, but this group is heterogeneous and likely reflects more than one disorder.

Vascular endothelial growth factor (VEGF) is a complex signaling protein that supports vascular and neuronal function. Alzheimer's disease (AD) -neuropathological hallmarks interfere with VEGF signaling and modify previously detected positive associations between cerebral spinal fluid (CSF) VEGF and cognition and hippocampal volume. However, it remains unknown 1) whether regional relationships between VEGF and glucose metabolism and cortical thinning exist, and 2) whether AD-neuropathological hallmarks (CSF Aβ, t-tau, p-tau) also modify these relationships. We addressed this in 310 Alzheimer's Disease Neuroimaging Initiative (ADNI) participants (92 cognitively normal, 149 mild cognitive impairment, 69 AD; 215 CSF Aβ+, 95 CSF Aβ-) with regional cortical thickness and cognition measurements and 158 participants with FDG-PET. In Aβ + participants (CSF Aβ42 ≤ 192 pg/mL), higher CSF VEGF levels were associated with greater FDG-PET signal in the inferior parietal, and middle and inferior temporal cortices. Abnormal CSF amyloid and tau levels strengthened the positive association between VEGF and regional FDG-PET indices. VEGF also had both direct associations with semantic memory, as well as indirect associations mediated by regional FDG-PET signal to cognition.

The relationship between magnetic resonance imaging (MRI)-visible enlarged perivascular spaces (EPVS) and Aβ and tau deposition is poorly investigated in cognitively normal older population. In our study, a total of 106 cognitively normal older subjects from the Alzheimer's Disease Neuroimaging Initiative database were included. All the subjects underwent brain MRI, florbetapir positron emission tomography (PET), and flortaucipir PET examinations. EPVS were rated on MRI using a 5-point scale in the basal ganglia (BG-EPVS) and the centrum semiovale (CSO-EPVS). Our study revealed that 43 subjects had high-degree BG-EPVS (degree >1) and 58 subjects had high-degree CSO-EPVS (degree >1). In logistic regression, high degree of BG-EPVS was associated with age (odds ratio [OR]: 1.08, 95% confidence interval [CI]: 1.01-1.16) and severe deep white matter hyperintensity (OR: 2.67, 95% CI: 1.12-6.35). High degree of CSO-EPVS was associated with flortaucipir PET positivity (OR: 2.24, 95% CI: 1.02-4.93). In conclusion, high degree of CSO-EPVS was associated with tau deposition in the brain, whereas high degree of BG-EPVS was associated with age and severe deep white matter hyperintensity, a marker of small vessel disease.

Some studies suggest that angiotensin II type 1 receptor blockers (ARBs) may protect against memory decline more than angiotensin-converting enzyme inhibitors (ACE-Is), but few have examined possible mechanisms. We assessed longitudinal differences between ARB versus ACE-I users in global and sub-regional amyloid-β accumulation by 18F-florbetapir. In cognitively normal older adults (n= 142), propensity-weighted linear mixed-effects models showed that ARB versus ACE-I use was associated with slower amyloid-β accumulation in the cortex, and specifically in the caudal anterior cingulate and precuneus, and in the precentral and postcentral gyri. In amyloid-positive participants with Alzheimer’s disease dementia or mild cognitive impairment (n = 169), ARB versus ACE-I use was not associated with different rates of amyloid-β accumulation. Apolipoprotein E e4 carrier status explained some heterogeneity in the different rates of amyloid-β accumulation between users of ARBs versus ACE-Is in the study. Replicative studies and clinical trials are warranted to confirm potential benefits of ARBs on rates of amyloid-β accumulation in the contexts of Alzheimer’s disease prevention and treatment.

BACKGROUND: Amyloid-β (Aβ) likely plays a primary role in Alzheimer’s disease pathogenesis, but longitudinal Aβ, tau, and neurodegeneration (A/T/N) measurements in the same individuals have rarely been examined to verify the temporal dynamics of these biomarkers. METHODS: In this study, we investigated the temporal ordering of Aβ, tau, and neurodegeneration using longitudinal biomarkers in nondemented elderly individuals. A total of 395 cognitively unimpaired individuals and 204 individuals with mild cognitive impairment (320 [53%] were female) were classified into 8 A±/T±/N± categories according to the abnormal (+)/normal (−) status of Aβ ((18)F-florbetapir or (18)F-florbetaben) positron emission tomography (PET), (18)F-flortaucipir PET, and adjusted hippocampal volume (aHCV). Follow-up Aβ PET, tau PET, and aHCV measurements at 0.6 to 4.1 years were available for 35% to 63% of the sample. Baseline Aβ, tau, and aHCV were compared between different A/T/N profiles. We investigated the associations of baseline and longitudinal Aβ, tau, and neurodegeneration in relation to one another continuously. RESULTS: Among T− participants, tau was higher for A+/T−/N− individuals compared with the A−/T−/N− group (p = .02). Among N− participants, neurodegeneration was worse among A+/T+/N− individuals compared with the A−/T−/N− group (p = .001). High baseline Aβ was associated (p < .001) with subsequent tau increase and high baseline tau was associated (p = .002) with subsequent aHCV decrease, whereas high tau and low aHCV at baseline were not associated with subsequent Aβ increase. CONCLUSIONS: These findings define a sequence of pathological events in Alzheimer’s disease that support a current model of Alzheimer’s disease pathogenesis in which Aβ appears early, followed by deposition of abnormal tau aggregates and subsequent neurodegeneration.

The relationship between depression and Alzheimer's disease (AD) is complex and still not well understood. We aimed to examine the roles of the AD core pathologies in modulating the associations of minimal depressive symptoms (MDSs) with cognitive impairments.A total of 721 participants who had measures of cognition, depressive symptoms, and cerebrospinal fluid AD biomarkers were included from the CABLE (Chinese Alzheimer's Biomarker and LifestylE) study. Causal mediation analyses with 10,000 bootstrapped iterations were conducted to explore the mediation effects of AD pathologies on cognition. The ADNI (Alzheimer's Disease Neuroimaging Initiative) was used 1) to replicate the mediation effects and 2) to examine the longitudinal relationships of MDSs with amyloid pathology and incident AD risk.In CABLE, MDSs were associated with poorer global cognition (p = .006) and higher amyloid burden as indicated by cerebrospinal fluid amyloid markers (p.0001). The influence of MDSs on cognition was partially mediated by amyloid pathology (a maximum of 85%). The mediation effects were replicated in 725 elderly persons without dementia (age, mean ± SD = 73.5 ± 6.9 years; 301 female subjects [42%]) in ADNI, such that the mediation percentage varied from 10% to 30% for general cognition, memory, and executive functions. Longitudinal analyses revealed a bidirectional relationship between MDSs and amyloid pathology (p = .01). MDSs were associated with 83% increased risk of developing AD dementia (hazard ratio = 1.83, p.01).Overall, amyloid pathology might partially mediate and magnify the influences of MDSs on cognitive impairments and AD risk.

Vascular contributions to early cognitive decline are increasingly recognized, prompting further investigation into the nature of related changes in perivascular spaces (PVS). Using magnetic resonance imaging, we show that, compared to a cognitively normal sample, individuals with early cognitive dysfunction have altered PVS presence and distribution, irrespective of Amyloid-β. Surprisingly, we noted lower PVS presence in the anterosuperior medial temporal lobe (1.29 times lower PVS volume fraction in cognitively impaired individuals compared with those unimpaired, p < 0.0001), which was associated with neurofibrillary tau tangle deposition in the right entorhinal cortex (beta (SE) = −0.98 (0.4); p = 0.014), one of the hallmarks of early Alzheimer’s disease pathology. We also observed higher PVS volume fraction in centrum semi-ovale of the white matter, but only in female participants (1.47 times higher PVS volume fraction in cognitively impaired individuals compared with those unimpaired, p = 0.0011). Furthermore, we observed PVS changes in participants with history of hypertension (higher PVS volume fraction in the white matter and lower PVS volume fraction in the anterosuperior medial temporal lobe). Our results suggest that anatomically specific alteration of the PVS is an early neuroimaging feature of cognitive impairment in aging adults, which is differentially manifested in female.

Neocortical Aβ-amyloid deposition, one of the hallmark pathologic features of Alzheimer's disease (AD), begins decades prior to the presence of clinical symptoms. As clinical trials move to secondary and even primary prevention, understanding the rates of neocortical Aβ-amyloid deposition and the age at which Aβ-amyloid deposition becomes abnormal is crucial for optimizing the timing of these trials. As APOE-ε4 carriage is thought to modulate the age of clinical onset, it is also important to understand the impact of APOE-ε4 carriage on the age at which the neocortical Aβ-amyloid deposition becomes abnormal. Here, we show that, for 455 participants with over 3 years of follow-up, abnormal levels of neocortical Aβ-amyloid were reached on average at age 72 (66.5-77.1). The APOE-ε4 carriers reached abnormal levels earlier at age 63 (59.6-70.3); however, noncarriers reached the threshold later at age 78 (76.1-84.4). No differences in the rates of deposition were observed between APOE-ε4 carriers and noncarriers after abnormal Aβ-amyloid levels had been reached. These results suggest that primary and secondary prevention trials, looking to recruit at the earliest stages of disease, should target APOE-ε4 carriers between the ages of 60 and 66 and noncarriers between the ages of 76 and 84.

In the present study, we tested the hypothesis that higher amyloid-beta (Aβ) burden at baseline is associated with greater longitudinal decline in body mass index (BMI) in clinically normal adults. Participants from the Harvard Aging Brain Study (n = 312) and the Alzheimer's Disease Neuroimaging Initiative (n = 336) underwent Aβ positron emission tomography at baseline. BMI was assessed longitudinally over a median of >4 years. Linear mixed models showed that higher baseline Aβ burden was significantly associated with greater decline in BMI in both the Harvard Aging Brain Study (t = −1.93; p = 0.05) and Alzheimer's Disease Neuroimaging Initiative cohorts (t = −2.54; p = 0.01), after adjusting for covariates, including cognitive performance and depressive symptoms. In addition, the association of Aβ burden with longitudinal decline in BMI persisted in both cohorts after excluding participants with diabetes/endocrine disturbances and participants classified as underweight or obese (BMI 30). These findings suggest that decline in BMI in clinically normal adults may be an early manifestation related to cerebral amyloidosis that precedes objective cognitive impairment. Therefore, unintentional BMI decline in otherwise healthy individuals might alert clinicians to increased risk of Alzheimer's disease.

HMG-CoA reductase inhibitors (statins) have been proposed to both positively and negatively alter the cognitive trajectory of older adults. To further investigate the disparate findings (provided by the FDA black box warning, adverse event reports, case-control studies, observational studies, and randomized controlled trials) that have led to such proposals, this study explored the association between statin use and longitudinal cognitive change, as well as time to diagnostic conversion, using prospective longitudinal (i.e., 24 months follow-up) cohort data from the Alzheimer’s Disease Neuroimaging Initiative (ADNI). No associations between statin use and cognitive change were found for participants with cognitively normal status, late mild cognitive impairment, or dementia due to Alzheimer’s disease. However, there was an association between statin use and slower decline on a memory composite in participants with early mild cognitive impairment. Lastly, results did not support an association between statin use and conversion to a more severe diagnostic category for any diagnostic group. Thus, these findings do not provide support for the FDA black box warning that claims statin use causes cognitive deficits, but rather, suggest that additional randomized trials investigating statin use in participants in the prodromal early mild cognitive impairment stage of Alzheimer’s disease may be warranted. OBJECTIVE: To investigate associations between statin use and cognitive change, as well as diagnostic conversion, in individuals with cognitively normal (CN) status, mild cognitive impairment (MCI), and dementia due to Alzheimer’s disease (AD-dementia). METHODS: A multi-center cohort study with 1629 adults 48 to 91 years-old with CN status, early MCI (EMCI), late MCI (LMCI), or AD-dementia at baseline followed prospectively for 24 months. Statin use was assessed at baseline, and cognition was measured over time with a composite memory score, a composite executive function score, and a global cognition score (Alzheimer’s Disease Assessment Scale). Conversion to a more impaired diagnostic category was determined by clinician assessment. Repeated measures linear mixed effects models were used to evaluate associations between statin use and change in cognition over time. Cox proportional hazards models were used to evaluate associations between statin use and time to diagnostic conversion. All models were stratified by baseline diagnostic group. RESULTS: Statin use was not associated with change in cognitive measures for CN, LMCI, or AD-dementia participants. Among EMCI participants, statin use was associated with a significantly slower rate of decline on the memory composite, but no other cognitive measure. Statin use was not associated with time to conversion for any diagnostic group. CONCLUSIONS: This study did not support an association between statin use and diagnostic conversion but suggested a possible association between statin use and cognitive change in EMCI. Additional randomized clinical trials of statins may be warranted in the prodromal EMCI stage of AD.

Background Stage 1 of the National Institute on Aging–Alzheimer’s Association’s proposed Alzheimer’s disease continuum is defined as amyloid-β (Aβ) positive but cognitively normal. Identifying at-risk individuals before Aβ reaches pathological levels could have great benefits for early intervention. Although Aβ levels become abnormal long before severe cognitive impairments appear, increasing evidence suggests that subtle cognitive changes may begin early, potentially before Aβ surpasses the threshold for abnormality. We examined whether baseline cognitive performance would predict progression from normal to abnormal levels of Aβ. Methods We examined the association of baseline cognitive composites (Preclinical Alzheimer Cognitive Composite, Alzheimer’s Disease Neuroimaging Initiative (ADNI) memory factor composite) with progression to Aβ positivity in 292 nondemented, Aβ-negative ADNI participants. Additional analyses included continuous cerebrospinal fluid biomarker levels to examine the effects of subthreshold pathology. Results Forty participants progressed to Aβ positivity during follow-up. Poorer baseline performance on both cognitive measures was significantly associated with increased odds of progression. More abnormal levels of baseline cerebrospinal fluid phosphorylated tau and subthreshold Aβ were associated with increased odds of progression to Aβ positivity. Nevertheless, baseline ADNI memory factor composite performance predicted progression even after controlling for baseline biomarker levels and APOE genotype (Preclinical Alzheimer Cognitive Composite was trend level). Survival analyses were largely consistent: controlling for baseline biomarker levels, baseline Preclinical Alzheimer Cognitive Composite still significantly predicted progression time to Aβ positivity (ADNI memory factor composite was trend level). Conclusions The possibility of intervening before Aβ reaches pathological levels is of obvious benefit. Low-cost, noninvasive cognitive measures can be informative for determining who is likely to progress to Aβ positivity, even after accounting for baseline subthreshold biomarker levels.

Brain age prediction is a machine learning method that estimates an individual’s chronological age from their neuroimaging scans. Brain age indicates whether an individual’s brain appears “older” than age-matched healthy peers, suggesting that they may have experienced a higher cumulative exposure to brain insults or were more impacted by those pathological insults. However, contemporary brain age models include older participants with amyloid pathology in their training sets and thus may be confounded when studying Alzheimer’s disease (AD). We showed that amyloid status is a critical feature for brain age prediction models. We trained a model on T1-weighted MRI images participants without amyloid pathology. MRI data were processed to estimate gray matter density voxel-wise, which were then used to predict chronological age. Our model performed accurately comparable to previous models. Notably, we demonstrated more significant differences between AD diagnostic groups than other models. Additionally, our model was able to delineate significant differences in brain age relative to chronological age between cognitively normal individuals with and without amyloid. Incorporation of amyloid status in brain age prediction models ultimately improves the utility of brain age as a biomarker for AD.

It is unclear whether episodic memory is an appropriate descriptor of the cognitive continuum in mild cognitive impairment (MCI). Here, we investigated the ability of episodic memory to track cognitive changes in patients with MCI with biomarker evidence of Alzheimer's disease (AD). We examined 387 MCI amyloid-positive subjects, cognitively staged as "early" or "late" on the basis of episodic memory impairment. Cross-sectional and longitudinal comparisons between these 2 groups were performed for each amyloid, tau, and neurodegeneration (AT(N)) profile. Cross-sectional analyses indicate that "early" MCI represents a transitional phase between normal cognition and "late" MCI in the AD biomarker pathway. After adjusting by confounders and levels of A, T, and (N), "late" MCI progressed significantly faster than "early" MCI only in profiles with both abnormal amyloid and tau markers (A+T+(N)- p < 0.05, A+T+(N)+ p < 0.001). Episodic memory staging is useful for describing symptoms in prodromal AD and complements the AT(N) profiles. Our findings might have implications for the Numeric Clinical staging scheme of the National Institute on Aging and Alzheimer's Association research framework.

A key question for the design of clinical trials for Alzheimer's disease (AD) is whether the timing of conversion from mild cognitive impairment (MCI) to AD can be predicted. This is also an important question for the clinical management of MCI. This study aims to address this question by exploring the contribution of baseline brain volume and annual volume change, using Cox regression, in predicting the time to conversion. Individuals with MCI, who converted to AD (n = 198), reverted to normal (n = 38), or remained stable (n = 96) for at least five years, were included in this study. The results revealed that the volumes of all the brain areas considered were predictive of the time to conversion from MCI to AD. Annual change in volume was also predictive of the time to conversion but only when initial volumes were above a certain threshold. This is important because it suggests that reduction in atrophy rate, which is the outcome of some clinical trials, is not inevitably associated with delay in conversion from MCI to AD.

Functional connectivity pattern altered of default mode network (DMN) is gaining more attention as a potential noninvasive biomarker to diagnose incipient Alzheimer's disease. However, the changed functional connectivity except for DMN, the longitudinal changes in executive control network (ECN) and frontoparietal network (FPN) also has attracted wide interest. Moreover, AD-related functional connectivity abnormalities within the DMN are well replicated research, but the (increased/decreased and reduced) functional connectivity in ECN and FPN weren't receive adequate attention. To address the above issues, in this paper, we adopt sparse inverse covariance estimation (SICE) approach to investigate the changed functional connectivity of ECN and FPN on the ADNI2 dataset. Our experimental results indicate the left superior frontal gyrus (SFGmed.L) and left thalamus (THA.L) regions of ECN has shown increased functional connectivity, the left anterior cingulate (ACG.L) region of ECN has shown decreased functional connectivity. The Superior Parietal Gyrus (SPG) regions and left paracentral lobule (PCL.L) of FPN has shown increased functional connectivity, the left supramarginal gyrus (SMG.L) regions has shown decreased functional connectivity in AD patients. On the other hand, the ACG.L regions in ECN, SMG.L and left inferior parietal (IPL.L) in FPN have shown significantly reduced functional connectivity. These results demonstrate that increased/decreased functional connectivity and reduced functional connectivity not only within DMN, but also associated with ECN and FPN. It also suggest that AD is associated with the characteristics of large-scale functional networks, and these changed functional connectivity possibly as a potential noninvasive biomarker to diagnose incipient Alzheimer's disease.

Empirical imaging biomarkers such as the level of the regional pathological burden are widely used to measure the risk of developing neurodegenerative diseases such as Alzheimer's disease (AD). However, ample evidence shows that the brain network (wirings of white matter fibers) plays a vital role in the progression of AD, where neuropathological burdens often propagate across the brain network in a prion-like manner. In this context, characterizing the spreading pathway of AD-related neuropathological events sheds new light on understanding the heterogeneity of pathophysiological mechanisms in AD. In this work, we propose a manifold-based harmonic network analysis approach to explore a novel imaging biomarker in the form of the AD propagation pattern, which eventually allows us to identify the AD-related spreading pathways of neuropathological events throughout the brain. The backbone of this new imaging biomarker is a set of region-adaptive harmonic wavelets that represent the common network topology across individuals. We conceptualize that the individual's brain network and its associated pathology pattern form a unique system, which vibrates as do all natural objects in the universe. Thus, we can computationally excite such a brain system using selected harmonic wavelets that match the system's resonance frequency, where the resulting oscillatory wave manifests the system-level propagation pattern of neuropathological events across the brain network. We evaluate the statistical power of our harmonic network analysis approach on large-scale neuroimaging data from ADNI. Compared with the other empirical biomarkers, our harmonic wavelets not only yield a new imaging biomarker to potentially predict the cognitive decline in the early stage but also offer a new window to capture the in-vivo spreading pathways of neuropathological burden with a rigorous mathematics insight.

Educational attainment is widely accepted as a cognitive reserve variable. However, few studies have demonstrated that education statistically moderates the effects of pathology on cognition. Here, we explored this issue in a sample of 441 Alzheimer's disease (AD) and mild cognitive impairment participants from the Alzheimer's Disease Neuroimaging Initiative cohort who had AD markers (Aβ42, tau, structural brain volumes) at baseline and underwent cognitive testing at baseline and at 6-month, 12-month, and 24-month time points. An AD-related biomarker (atrophy/pathology) composite at baseline was developed using stepwise backward linear regression. Potential moderation effects of education on the relationship between AD biomarkers and cognition were explored using linear mixed models. Education was positively correlated with cognition, and biomarkers were negatively correlated with cognition, across domains and diagnostic groups. However, education generally did not moderate the effects of biomarkers on baseline or longitudinal cognition. Our results do not support the hypothesis that education protects cognitive function from brain pathology in the Alzheimer's Disease Neuroimaging Initiative cohort, questioning its accepted status as a reserve variable.

We evaluated the associations of subjective (self-reported everyday cognition [ECog]) and objective cognitive measures with regional amyloid-β (Aβ) and tau accumulation in 86 clinically normal elderly subjects from the Alzheimer's Disease Neuroimaging Initiative. Regression analyses were conducted to identify whether individual ECog domains (Memory, Language, Organization, Planning, Visuospatial, and Divided Attention) were equally or differentially associated with regional [18F]florbetapir and [18F]flortaucipir uptake and how these associations compared to those obtained with objective cognitive measures. A texture analysis, the weighted 2-point correlation, was used as an additional approach for estimating the whole-brain tau burden without positron emission tomography intensity normalization. Although the strongest models for ECog domains included either tau (planning and visuospatial) or Aβ (memory and organization), the strongest models for all objective measures included Aβ. In Aβ-negative participants, the strongest models for all ECog domains of executive functioning included tau. Our results indicate differential associations of individual subjective cognitive domains with Aβ and tau in clinically normal adults. Detailed characterization of ECog may render a valuable prescreening tool for pathological prediction.

There is a pressing need to identify individuals at high risk of conversion from mild cognitive impairment (MCI) to Alzheimer's disease (AD) based on available repeated cognitive measures in primary care. Using data from the Alzheimer's Disease Neuroimaging Initiative (ADNI), we applied a joint latent class mixed model (JLCM) to derive a 3-class solution: low risk (72.65%), medium risk (20.41%) and high risk (6.94%). In the low-risk group, individuals with lower daily activity and ApoEε4 carriers were at greater risk of conversion from MCI to AD. In the medium-risk group, being female, single, and an ApoEε4 carrier increased risk of conversion to AD. In the high-risk group, individuals with lower education level and single individuals were at greater risk of conversion to AD. Individual dynamic prediction for conversion from MCI to AD after 10 years was derived. Accurate identification of conversion from MCI to AD contributes to earlier close monitoring, appropriate management, and targeted interventions. Thereby, it can reduce avoidable hospitalizations for the high-risk MCI population. Moreover, it can avoid expensive follow-up tests that may provoke unnecessary anxiety for low-risk individuals and their families.

Cholecystokinin (CCK) is a satiety hormone that is highly expressed in brain regions like the hippocampus. CCK is integral for maintaining or enhancing memory and thus may be a useful marker of cognitive and neural integrity in participants with normal cognition, mild cognitive impairment, and Alzheimer's disease (AD). Cerebrospinal fluid (CSF) CCK levels were examined in 287 subjects from the Alzheimer's Disease Neuroimaging Initiative. Linear or voxelwise regression was used to examine associations between CCK, regional gray matter, CSF AD biomarkers, and cognitive outcomes. Briefly, higher CCK was related to a decreased likelihood of having mild cognitive impairment or AD, better global and memory scores, and more gray matter volume primarily spanning posterior cingulate cortex, parahippocampal gyrus, and medial prefrontal cortex. CSF CCK was also strongly related to higher CSF total tau (R2 = 0.342) and p-tau-181 (R2 = 0.256) but not Aβ1-42. Tau levels partially mediated CCK and cognition associations. In conclusion, CCK levels may reflect compensatory protection as AD pathology progresses.

Alzheimer's disease (AD) is an irreversible neurodegenerative disorder with progressive impairment of memory and other mental functions. Magnetic resonance images (MRI) have been widely used as an important imaging modality of brain for AD diagnosis and monitoring the disease progression. The longitudinal analysis of sequential MRIs is important to model and measure the progression of the disease along the time axis for more accurate diagnosis. Most existing methods extracted the features capturing the morphological abnormalities of brain and their longitudinal changes using MRIs and then designed a classifier to discriminate different groups. However, these methods have several limitations. First, since the feature extraction and classifier model are independent, the extracted features may not capture the full characteristics of brain abnormalities related to AD. Second, longitudinal MR images may be missing at some time points for some subjects, which results in difficulties for extraction of consistent features for longitudinal analysis. In this paper, we present a classification framework based on combination of convolutional and recurrent neural networks for longitudinal analysis of structural MR images in AD diagnosis. First, Convolutional Neural Networks (CNN) is constructed to learn the spatial features of MR images for the classification task. After that, recurrent Neural Networks (RNN) with cascaded three bidirectional gated recurrent units (BGRU) layers is constructed on the outputs of CNN at multiple time points for extracting the longitudinal features for AD classification. Instead of independently performing feature extraction and classifier training, the proposed method jointly learns the spatial and longitudinal features and disease classifier, which can achieve optimal performance. In addition, the proposed method can model the longitudinal analysis using RNN from the imaging data at various time points. Our method is evaluated with the longitudinal T1-weighted MR images of 830 participants including 198 AD, 403 mild cognitive impairment (MCI), and 229 normal controls (NC) subjects from Alzheimer's Disease Neuroimaging Initiative (ADNI) database. Experimental results show that the proposed method achieves classification accuracy of 91.33% for AD vs. NC and 71.71% for pMCI vs. sMCI, demonstrating the promising performance for longitudinal MR image analysis.

The CD40 pathway has been implicated in microglial activation, which is considered as a key factor in the pathogenesis of Alzheimer’s disease (AD). However, the association of CSF CD40 and synaptic degeneration in living human is not clear. A total of 294 subjects with different severities of cognitive impairments were included in this study: 84 participants with normal cognition, 143 patients with mild cognitive impairment (MCI) and 67 patients with mild AD. Levels of CD40 in CSF were compared among the three groups. Further, several linear regression models were conducted to explore the associations of CSF CD40 and neurogranin levels (reflecting synaptic degeneration) when controlling for age, gender, educational attainment, APOE4 genotype, clinical diagnosis, CSF Aβ42 and tau proteins. We found that CSF CD40 levels were significantly decreased in patients with mild AD compared with healthy controls and MCI patients (control vs. AD, p = 0.0026; MCI vs. AD, p = 0.0268). However, there were no significant differences in CSF CD40 levels between controls and patients with MCI (p = 0.37). In addition, CSF CD40 levels were associated with neurogranin in the pooled sample when controlling for age, gender, educational attainment, APOE4 genotype and diagnosis. In summary, our findings support the notion that the CD40 pathway may contribute to an important mechanism underlying synaptic degeneration in AD.

TREM2 was suggested to be an important regulator of microglia during neurodegeneration, but previous studies report conflicting results in relation to soluble TREM2 (sTREM2) in cerebrospinal fluid (CSF) when using clinical criteria to classify Alzheimer's disease (AD). The present study explores sTREM2 CSF levels and their associations with other biomarkers and cognitive measures in a prospective AD cohort. Based on the available CSF biomarker information, 497 subjects were classified according to the 2018 National Institute on Aging-Alzheimer's Association research framework guidelines, which group biomarkers into those of amyloid-β deposition, tau pathology, and neurodegeneration. CSF sTREM2 concentrations were associated with markers of neurodegeneration and fibrillar tau pathology, but not amyloidosis; sTREM2 concentrations were increased in total tau-positive versus -negative individuals; sTREM2 was not related to cognitive and other biomarker changes over time; and sTREM2 concentrations increased over time in total tau-positive versus -negative individuals with AD pathophysiology. The present study provides evidence in support of sTREM2 in CSF as a marker of neuroinflammation across the spectrum of early clinical AD. sTREM2 is linked to neuronal injury and may therefore offer complementary information relevant for diagnostic purposes and novel treatment approaches targeting the immune system in AD.

Longitudinal MRI studies are often subjected to mid-study scanner changes, which may alter image characteristics such as contrast, signal-to-noise ratio, contrast-to-noise ratio, intensity non-uniformity and geometric distortion. Measuring brain volume loss under these conditions can render the results potentially unreliable across the timepoint of the change. Estimating and accounting for this effect may improve the reliability of estimates of brain atrophy rates.We analyzed 237 subjects who were scanned at 1.5 T for the Alzheimer's Disease Neuroimaging Initiative (ADNI) study and were subject to intra-vendor or inter-vendor scanner changes during follow-up (up to 8 years). Sixty-three subjects scanned on GE Signa HDx and HDxt platforms were also subject to a T1-weighted sequence change from Magnetization Prepared Rapid Gradient Echo (MP-RAGE) to Fast Spoiled Gradient Echo with IR Preparation (IR-FSPGR), as part of the transition from ADNI-1 to ADNI-2/GO. Two-timepoint percentage brain volume changes (PBVCs) between the baseline "screening" and the follow-up scans were calculated using SIENA. A linear mixed-effects model with subject-specific random slopes and intercepts was applied to estimate the fixed effects of scanner hardware changes on the PBVC measures. The same model also included a term to estimate the fixed effects of the T1-weighted sequence change.Different hardware upgrade or change combinations led to different offsets in the PBVC (SE; p): Philips Intera to Siemens Avanto, -1.81% (0.30; p 0.0001); GE Genesis Signa to Philips Intera, 0.99% (0.47, p = 0.042); GE Signa Excite to Signa HDx, 0.33% (0.095, p = 0.0005); GE Signa Excite to Signa HDxt, -0.023% (0.23, p = 0.92); GE Signa Excite to Signa HDx to Signa HDxt, 0.25% (0.095, p = 0.010) and 0.27% (0.16, p = 0.098), respectively; GE Signa HDx to Signa HDxt, -0.24% (0.25, p = 0.34); Siemens Symphony to Symphony TIM, -0.39% (0.16; p = 0.019). The sequence change from MP-RAGE to IR-SPGR was associated with an average -1.63% (0.12; p 0.0001) change.Inter-vendor scanner changes generally led to greater effects on PBVC measurements than did intra-vendor scanner upgrades. The effect of T1-weighted sequence change was comparable to that of the inter-vendor scanner changes. Inclusion of the corrective fixed-effects terms for the scanner hardware and T1-weighted sequence changes yielded better model goodness-of-fits, and thus, potentially more reliable estimates of whole-brain atrophy rates.

The most prevalent and common type of dementia is Alzheimer's disease (AD). However, it is notable that very few people who are suffering from AD are diagnosed correctly and in a timely manner. The definite cause and cure of the disease are still unavailable. The symptoms might be more manageable and its treatment can be more effective, when the impairment is still at an earlier stage or at MCI (mild cognitive impairment). AD can be clinically diagnosed by physical and neurological examination, so there is an need for developing better and efficient diagnostic tools for AD. In recent years, content-based image retrieval (CBIR) systems have been widely researched and applied in many medical applications. Combining an automated image classification system and the radiologist's professional knowledge, to increase the accuracy of prediction and diagnosis, were the main motives. In this paper, a multistage classifier using machine learning, including Naive Bayes classifier, support vector machine (SVM), and K-nearest neighbor (KNN), was used to classify Alzheimer's disease more acceptably and efficiently. For this, MRI (Magnetic resonance imaging) scans were processed by FreeSurfer, a powerful software tool suitable for processing and normalizing brain MRI images. We also applied a feature selection technique - PSO (particle swarm optimization) to many feature vectors in order to obtain the best features that represent the salient characteristics of AD. The results of the proposed method outperform individual techniques in a benchmark database provided by the Alzheimer's Disease Neuroimaging Institute (ADNI). Keywords: Alzheimer's disease, Machine learning, Content-based image retrieval, Multistage classifier, PSO, Structural MRI, SVM, K-NN

Mild cognitive impairment (MCI) of the amnestic type is considered to be a transitionary stage between healthy aging and Alzheimer’s disease (AD). Previous studies have demonstrated that intrinsic functional connectivity of the default network (DN) is altered in normal aging and AD and impacts both within and between network connectivity. While changes within the DN have been reported in MCI, it remains uncertain how interactions with other large-scale brain networks are altered in this prodromal stage of AD. We investigated within and between network connectivity in healthy older adults (HOAs) and older adults with MCI across three canonical brain networks: DN, dorsal attention network, and frontoparietal control network. We also assessed how patterns of functional connectivity among the three networks predicted cognitive status and age using multivariate partial least squares. A total of 91 MCI and 71 HOA restingstate scans were analyzed from the Alzheimer’s Disease Neuroimaging Initiative. There were three key findings. First, a circumscribed pattern of greater between network and interhemispheric connectivity was associated with higher cognitive status in HOAs. Second, for individuals with MCI, cognitive status was positively associated with a more distributed, less differentiated pattern of intrinsic functional connectivity across the three networks. Finally, greater within network functional connectivity was positively associated with cognitive status for HOAs irrespective of age, however this compensation-like effect diminished with increasing age for MCI participants. While reliable differences between healthy aging and MCI in the intrinsic network architecture of the brain are apparent, these differences emerge as shifting associations between network interactivity, cognitive functioning and age.

The association of vascular endothelial growth factor (VEGF) levels in CSF and cerebral glucose metabolism across the Alzheimer’s disease (AD) spectrum is unclear. CSF VEGF levels were cross-sectionally related to cerebral glucose metabolism, as measured by 18F-fluorodeoxyglucose positron emission tomography (FDG-PET), using linear regression models. We found that VEGF levels were associated with cerebral glucose metabolism in patients with mild cognitive impairment (MCI) and AD, but not in cognitively normal older adults. Our data indicated that VEGF may play an important role in cerebral glucose metabolism.

Background Optimized abnormalities of individual brain network may allow earlier detection of mild cognitive impairment (MCI) and accurate prediction of its conversion to Alzheimer’s disease (AD). Currently, most studies constructed individual networks based on region-to-region correlation without employing multi-region information. In order to develop the potential discriminative power of network and provide supportive evidence for feasibility of individual metabolic network study, we propose a new approach to extract features from network with indirect relation based on 18F-FDG PET data. New Method Direct relation based individual network is first constructed using Gaussian kernel function. After that, the lattice-close-degree in fuzzy mathematics is applied to reflect region-to-region indirect relation using the direct relations of regions and their common neighbors. The proposed approach has been evaluated on 199 MCI subjects and 166 normal controls (NC) using SVM classifier. Results The indirect relation based network features significantly promote classification performance in separating MCI from normal controls (NC) as well as MCI converters from non-converters. Specially, further improvements can be obtained by combining indirect relation features with ADAS-cog scores. Moreover, the discriminative regions we found are consistent with previous studies, indicating the efficacy of our constructed network in identifying correct biomarkers for diagnosing MCI and predicting its conversion. Comparison with Existing Method(s) More accurate MCI identification of PET data can be achieved by features of network with indirect relation. Conclusions This work provides a new way to investigate brain network from metabolic perspective for accurate identification of MCI.

Disease heterogeneity is a significant obstacle to understanding pathological processes and delivering precision diagnostics and treatment. Clustering methods have gained popularity for stratifying patients into subpopulations (i.e., subtypes) of brain diseases using imaging data. However, unsupervised clustering approaches are often confounded by anatomical and functional variations not related to a disease or pathology of interest. Semi-supervised clustering techniques have been proposed to overcome this and, therefore, capture disease-specific patterns more effectively. An additional limitation of both unsupervised and semi-supervised conventional machine learning methods is that they typically model, learn and infer from data using a basis of feature sets pre-defined at a fixed anatomical or functional scale (e.g., atlas-based regions of interest). Herein we propose a novel method, "Multi-scAle heteroGeneity analysIs and Clustering" (MAGIC), to depict the multi-scale presentation of disease heterogeneity, which builds on a previously proposed semi-supervised clustering method, HYDRA. It derives multi-scale and clinically interpretable feature representations and exploits a double-cyclic optimization procedure to effectively drive identification of inter-scale-consistent disease subtypes. More importantly, to understand the conditions under which the clustering model can estimate true heterogeneity related to diseases, we conducted extensive and systematic semi-simulated experiments to evaluate the proposed method on a sizeable healthy control sample from the UK Biobank (N = 4403). We then applied MAGIC to imaging data from Alzheimer's disease (ADNI, N = 1728) and schizophrenia (PHENOM, N = 1166) patients to demonstrate its potential and challenges in dissecting the neuroanatomical heterogeneity of common brain diseases. Taken together, we aim to provide guidance regarding when such analyses can succeed or should be taken with caution. The code of the proposed method is publicly available at https://github.com/anbai106/MAGIC.

Background Due to the low signal-to-noise-ratio (SNR) and unavoidable head motions, the pairwise subtraction perfusion signal extraction process in arterial spin labeling (ASL) perfusion MRI can produce extreme outliers. Comparison with existing methods We previously proposed an adaptive outlier cleaning (AOC) algorithm for ASL MRI. While it performed well even for clinical ASL data, two issues still exist. One is that if the reference is already dominated by noise, outlier cleaning using low correlation with the mean as a rejection criterion will actually reject the less noisy samples but keep the more noisy ones. The other is that it is sub-optimal to reject the entire outlier volumes without considering the quality of each constituent slices. To address both problems, a prior-guided and slice-wise AOC algorithm was proposed in this study. New Methods The reference of AOC was initiated to be a pseudo cerebral blood flow (CBF) map based on prior knowledge and outlier rejection was performed at each slice. ASL data from the ADNI database ( www.adni-info.org ) were used to validate the method. Image preprocessing was performed using ASLtbx. Results The proposed method outperformed the original AOC and SCORE in terms of higher SNR and test-retest stability of the resultant CBF maps. Conclusion ASL CBF can be substantially improved using prior-guided and slice-wise outlier rejection. The proposed method will benefit the ever since increasing ASL user community for both clinical and scientific brain research.

Alzheimer's disease (AD) is a severe neurodegenerative disorder characterized by loss of memory and reduction in cognitive functions due to progressive degeneration of neurons and their connections, eventually leading to death. In this paper, we consider the problem of simultaneously predicting several different cognitive scores associated with categorizing subjects as normal, mild cognitive impairment (MCI), or Alzheimer's disease (AD) in a multi-task learning framework using features extracted from brain images obtained from ADNI (Alzheimer's Disease Neuroimaging Initiative). To solve the problem, we present a multi-task sparse group lasso (MT-SGL) framework, which estimates sparse features coupled across tasks, and can work with loss functions associated with any Generalized Linear Models. Through comparisons with a variety of baseline models using multiple evaluation metrics, we illustrate the promising predictive performance of MT-SGL on ADNI along with its ability to identify brain regions more likely to help the characterization Alzheimer's disease progression.

Structural deterioration and volume loss of the hippocampal formation is observed in many diseases associated with memory decline. Paradoxically, glucose metabolism of the hippocampal formation can be increased at the same time. This might be a consequence of compensatory (beneficial) or maladaptive (detrimental) mechanisms. Aim of this study was to differentiate between compensation and maladaptation by analyzing the association between glucose metabolism in the hippocampal formation measured by positron emission tomography with the glucose analogue 18F-fluorodeoxyglucose and cognitive performance as characterized by the extended Consortium to Establish a Registry for Alzheimer's Disease test battery in a sample of 87 patients (81.8 ± 5.4 years) with mild cognitive impairment or mild dementia and varying etiological diagnoses. Glucose metabolism in the hippocampal formation was negatively correlated with the performance in several cognitive subdomains, most pronounced for verbal semantic fluency, independent of overall neuronal dysfunction, presence of clinical Alzheimer's disease, and overall cognitive performance. This finding provides evidence that increased glucose metabolism in the hippocampal formation of cognitively impaired patients indicates detrimental maladaptation rather than a beneficial compensatory reaction. Excess glucose metabolism in the hippocampal formation might be a useful therapeutic target in these patients.

Background Early diagnosis of Alzheimer's disease (AD) and its onset in subjects affected by mild cognitive impairment (MCI) based on structural MRI features is one of the most important open issues in neuroimaging. Accordingly, a scientific challenge has been promoted, on the international Kaggle platform, to assess the performance of different classification methods for prediction of MCI and its conversion to AD. New method This work presents a classification strategy based on Random Forest feature selection and Deep Neural Network classification using a mixed cohort including the four classes of classification problem, that is HC, AD, MCI and cMCI, to train the model. Moreover, we compare this approach with a novel classification strategy based on fuzzy logic learned on a mixed cohort including only HC and AD. Experiments A training set of 240 subjects and a test set including mixed cohort of 500 real and simulated subjects were used. The data included AD patients, MCI subjects converting to AD (cMCI), MCI subjects and healthy controls (HC). This work ranked third for overall accuracy (38.8%) over 19 participating teams. Comparison with existing method(s) The “International challenge for automated prediction of MCI from MRI data” hosted by the Kaggle platform has been promoted to validate different methodologies with a common set of data and evaluation procedures. Conclusion DNNs reach a classification accuracy significantly higher than other machine learning strategies; on the other hand, fuzzy logic is particularly accurate with cMCI, suggesting a combination of these approaches could lead to interesting future perspectives.

Background Previous publications indicate that Alzheimer's Disease (AD) related cortical atrophy may develop in asymmetric patterns, with accentuation of the left hemisphere. Since fluorodeoxyglucose positron emission tomography (FDG PET) measurements of the regional cerebral metabolic rate of glucose (rCMRgl) provide a sensitive and specific marker of neurodegenerative disease progression, we sought to investigate the longitudinal pattern of rCMRgl in amyloid-positive persons with mild cognitive impairment (MCI) and dementia, hypothesizing asymmetric declines of cerebral glucose metabolism. Methods Using florbetapir PET and cerebrospinal fluid (CSF) measures to define amyloid-β (Aβ) positivity, 40 Aβ negative (Aβ-) cognitively unimpaired controls (CU; 76 ± 5y), 76 Aβ positive (Aβ+) persons with MCI (76 ± 7y) and 51 Aβ + persons with probable AD dementia (75 ± 7y) from the AD Neuroimaging Initiative (ADNI) were included in this study with baseline and 2-year follow-up FDG PET scans. The degree of lateralization of longitudinal rCMRgl declines in subjects with Aβ + MCI and AD in comparison with Aβ- CU were statistically quantified via bootstrapped lateralization indices [(LI); range − 1 (right) to 1 (left)]. Results Compared to Aβ- CU, Aβ + MCI patients showed marked left hemispheric lateralization (LI: 0.78). In contrast, modest right hemispheric lateralization (LI: −0.33) of rCMRgl declines was found in Aβ + persons with probable AD dementia. Additional comparisons of Aβ + groups (i.e. MCI and probable AD dementia) consequently indicated right hemispheric lateralization (LI: −0.79) of stronger rCMRgl declines in dementia stages of AD. For all comparisons, voxel-based analyses confirmed significant (pFWE, Highlights • Lateralized cerebral glucose metabolism declines in Alzheimer's Disease. • Early stages show strong left-hemispheric lateralization. • Advanced stages show weak right-hemispheric lateralization.

Background The Trail Making Test (TMT) Part A (TMT-A) is a good measure of performance on cognitive processing speed. This study aimed to perform a genome-wide association study of TMT-A in Alzheimer's disease (AD). Methods A total of 757 individuals with TMT-A phenotypes and 620,901 single nucleotide polymorphisms (SNPs) were extracted from the Alzheimer's Disease Neuroimaging Initiative 1 (ADNI-1) cohort. AD related cognitive phenotypes include TMT-A, TMT-B, Functional Activities Questionnaire (FAQ), Clinical Dementia Rating Sum of Boxes (CDR-SB), and Alzheimer's Disease Assessment Scale–Cognitive Subscale 13 (ADAS13). Multivariable linear regression analysis of TMT-A was conducted using PLINK software. The most TMT-A associated gene was tested with Color Trails Test 1 Form A (CTTA), a culturally fair analog of the TMT-A. Functional annotation of SNPs was performed using the RegulomeDB and Genotype-Tissue Expression (GTEx) databases. Results The best signal with TMT-A was rs1108010 (p = 4.34 × 10−8) at 11p15.2 within INSC gene, which was also associated with TMT-B, FAQ, CDR-SB, and ADAS13 (p = 2.47 × 10−4, 8.56 × 10−3, 0.0127 and 0.0188, respectively). Furthermore, suggestive loci were identified such as FOXD2 and CLTA with TMT-A, GBP1/GBP3 with TMT-B, GRIK2 with FAQ, BAALC and CCDC146 with CDR-SB, BAALC and NKAIN2 with ADAS13. Additionally, the best SNP within INSC associated with CTTA was rs7931705 (p = 6.15 × 10−5). Several SNPs had significant eQTLs using GTEx. Conclusions We identified several genes/loci associated with TMT-A and AD related phenotypes. These findings offer the potential for new insights into the pathogenesis of cognitive function and Alzheimer's disease.

Alzheimer's disease (AD) is a neurodegenerative disease that afflicts millions of people worldwide. Early detection of AD is critical, as drug trials show a promising advantage to those patients with early diagnoses. In this study, magnetic resonance imaging (MRI) datasets from the Alzheimer's Disease Neuroimaging Initiative (ADNI) and The Open Access Series of Imaging Studies are used. Our method for performing the classification of AD is to combine a set of shearlet-based descriptors with deep features. A major challenge in classifying such MRI datasets is the high dimensionality of feature vectors because of the large number of slices of each MRI sample. Given the volumetric nature of the MRI data, we propose using the 3D shearlet transform (3D-ST), but we obtain the average of all directionalities, which reduces the dimensionality. On the other hand, we propose to leverage the capabilities of convolutional neural networks (CNN) to learn feature maps from stacked MRI slices, which generate a very compact feature vector for each MRI sample. The 3D-ST and CNN feature vectors are combined for the classification of AD. After the concatenation of the feature vectors, they are used to train a classifier. Alternatively, a custom CNN model is utilized, in which the descriptors are further processed end to end to obtain the classification model. Our experimental results show that the fusion of shearlet-based descriptors and deep features improves classification performance, especially on the ADNI dataset.

Genome-wide association analysis (GWAS) is a commonly used method to detect the potential biomarkers of Alzheimer's disease (AD). Most existing GWAS methods entail a high computational cost, disregard correlations among imaging data and correlations among genetic data, and ignore various associations between longitudinal imaging and genetic data. A novel GWAS method was proposed to identify potential AD biomarkers and address these problems. A network based on a gated recurrent unit was applied without imputing incomplete longitudinal imaging data to integrate the longitudinal data of variable lengths and extract an image representation. In this study, a modified diet network that can considerably reduce the number of parameters in the genetic network was proposed to perform GWAS between image representation and genetic data. Genetic representation can be extracted in this way. A link between genetic representation and AD was established to detect potential AD biomarkers. The proposed method was tested on a set of simulated data and a real AD dataset. Results of the simulated data showed that the proposed method can accurately detect relevant biomarkers. Moreover, the results of real AD dataset showed that the proposed method can detect some new risk-related genes of AD. Based on previous reports, no research has incorporated a deep-learning model into a GWAS framework to investigate the potential information on super-high-dimensional genetic data and longitudinal imaging data and create a link between imaging genetics and AD for detecting potential AD biomarkers. Therefore, the proposed method may provide new insights into the underlying pathological mechanism of AD.

We explored whether patients with Parkinson's disease dementia (PDD) show a distinct spatial metabolic pattern that characterizes cognitive deficits in addition to motor dysfunction. Eighteen patients with PDD underwent 3 separate positron emission tomography sessions with [ 18 F]fluorodeoxyglucose (for glucose metabolism), fluorinated N-3-fluoropropyl-2-beta-carboxymethoxy-3-beta-(4-iodophenyl) nortropane (for dopamine transporter density) and Pittsburgh compound-B (for beta-amyloid load). We confirmed in PDD versus normal controls, overall hypometabolism in the posterior and prefrontal brain regions accompanied with hypermetabolism in subcortical structures and the cerebellar vermis. A multivariate network analysis then revealed 3 metabolic patterns that are separately associated with cognitive performance ( p = 0.042), age ( p = 0.042), and motor symptom severity ( p = 0.039). The age-related pattern's association with aging was replicated in healthy controls ( p = 0.047) and patients with Alzheimer's disease ( p = 0.002). The cognition-related pattern's association with cognitive performance was observed, with a trend-level of correlation, in patients with dementia with Lewy bodies ( p = 0.084) but not in patients with Alzheimer's disease ( p = 0.974). We found no association with fluorinated N-3-fluoropropyl-2-beta-carboxymethoxy-3-beta-(4-iodophenyl) nortropane and Pittsburgh compound-B positron emission tomography with patients' cognitive performance.

Identifying individuals destined to develop Alzheimer's dementia within time frames acceptable for clinical trials constitutes an important challenge to design studies to test emerging disease-modifying therapies. Although amyloid-β protein is the core pathologic feature of Alzheimer's disease, biomarkers of neuronal degeneration are the only ones believed to provide satisfactory predictions of clinical progression within short time frames. Here, we propose a machine learning–based probabilistic method designed to assess the progression to dementia within 24 months, based on the regional information from a single amyloid positron emission tomography scan. Importantly, the proposed method was designed to overcome the inherent adverse imbalance proportions between stable and progressive mild cognitive impairment individuals within a short observation period. The novel algorithm obtained an accuracy of 84% and an under-receiver operating characteristic curve of 0.91, outperforming the existing algorithms using the same biomarker measures and previous studies using multiple biomarker modalities. With its high accuracy, this algorithm has immediate applications for population enrichment in clinical trials designed to test disease-modifying therapies aiming to mitigate the progression to Alzheimer's disease dementia.

Rationale and objectives Early prediction of incipient Alzheimer's disease (AD) dementia in individuals with mild cognitive impairment (MCI) is important for timely therapeutic intervention and identifying participants for clinical trials at greater risk of developing AD. Methods to predict incipient AD in MCI have mostly utilized cross-sectional data. Longitudinal data enables estimation of the rate of change of variables, which along with the variable levels have been shown to improve prediction power. While some efforts have already been made in this direction, all previous longitudinal studies have been based on observation periods longer than one year, hence limiting their practical utility. It remains to be seen if follow-up evaluations within shorter intervals can significantly improve the accuracy of prediction in this problem. Our aim was to determine the added value of incorporating 6-month longitudinal data for predicting progression from MCI to AD. Materials and methods Using 6-months longitudinal data from 247 participants with MCI, we trained two Random Forest classifiers to distinguish between progressive MCI (n = 162) and stable MCI (n = 85) cases. These models utilized structural MRI, neurocognitive assessments, and demographic information. The first model (cross-sectional) only used baseline data. The second model (longitudinal) used data from both baseline and a 6-month follow-up evaluation allowing the model to additionally incorporate biomarkers’ rate of change. Results The longitudinal model (AUC = 0.87; accuracy = 80.2%) performed significantly better (P Conclusion Short-term longitudinal assessments significantly enhance the performance of AD prediction models.

Background: The corpus callosum is the largest white matter tract in the human brain, involved in inter-hemispheric transfer and integration of lateralised visual, sensory-motor, language, and cognitive information. Microstructural alterations are implicated in ageing as well as various neurological conditions. \ud \ud New Method: Cross-sectional diffusion-weighted images of 107 healthy adults were used to create a linear regression model of the ageing corpus callosum and its sub-regions to evaluate the impact of analysis by sub-region, and to test for deviations from healthy ageing parameters in 28 subjects with mild cognitive impairment (MCI). Alterations in diffusion properties including fractional anisotropy, mean, radial and axial diffusivities were investigated as a function of age. \ud \ud Results: Changes in DTI parameters showed age-dependent regional differences, likely arising from axonal diameter variation across cross-sectional regions of interest in the corpus callosum. Patterns suggestive of degeneration with healthy ageing were observed in all regions. Diffusion parameters in sub-regions projecting to pre-motor, primary, and supplementary motor areas of the brain differed for MCI versus healthy controls, and MCI subjects were more likely than healthy controls to experience a reduction in motor skills. \ud \ud Comparison with Existing Methods: Statistical analyses of the corpus callosum by five manually-defined sub-regions, instead of a single manually-defined region of interest, revealed region-specific changes in microstructure in healthy ageing and MCI, and accounted for clinically-evaluated differences in motor skills between cohorts. \ud \ud Conclusion: This method will support future studies of corpus callosum, enabling identification and measurement of white matter changes that are undetectable with the single ROI approach.

Alzheimer's Disease (AD) is a chronic neurodegenerative disease without effective medications or supplemental treatments. Thus, predicting AD progression is crucial for clinical practice and medical research. Due to limited neuroimaging data, two-dimensional convolutional neural networks (2D CNNs) have been commonly adopted to differentiate among cognitively normal subjects (CN), people with mild cognitive impairment (MCI), and AD patients. Therefore, this paper proposes an ensemble learning (EL) architecture based on 2D CNNs, using a multi-model and multi-slice ensemble. First, the top 11 coronal slices of grey matter density maps for AD versus CN classifications were selected. Second, the discriminator of a generative adversarial network, VGG16, and ResNet50 were trained with the selected slices, and the majority voting scheme was used to merge the multi-slice decisions of each model. Afterwards, those three classifiers were used to construct an ensemble model. Multi-slice ensemble learning was designed to obtain spatial features, while multi-model integration reduced the prediction error rate. Finally, transfer learning was used in domain adaptation to refine those CNNs, moving them from working solely with AD versus CN classifications to being applicable to other tasks. This ensemble approach achieved accuracy values of 90.36%, 77.19%, and 72.36% when classifying AD versus CN, AD versus MCI, and MCI versus CN, respectively. Compared with other state-of-the-art 2D studies, the proposed approach provides an effective, accurate, automatic diagnosis along the AD continuum. This technique may enhance AD diagnostics when the sample size is limited.

Data-driven disease progression models have provided important insight into the timeline of brain changes inAD phenotypes. However, their utility in predicting the progression of pre-symptomatic AD in a populationbased setting has not yet been investigated. In this study, we investigated if the disease timelines constructedin a case-controlled setting, with subjects stratified according to APOE status, are generalizable to a populationbased cohort, and if progression along these disease timelines is predictive of AD. Seven volumetric biomarkersderived from structural MRI were considered. We estimated APOE-specific disease timelines of changes in thesebiomarkers using a recently proposed method called co-initialized discriminative event-based modeling (co-initDEBM). This method can also estimate a disease stage for new subjects by calculating their position along thedisease timelines. The model was trained and cross-validated on the Alzheimer’s Disease Neuroimaging Initiative(ADNI) dataset, and tested on the population-based Rotterdam Study (RS) cohort. We compared the diagnosticand prognostic value of the disease stage in the two cohorts. Furthermore, we investigated if the rate of change ofdisease stage in RS participants with longitudinal MRI data was predictive of AD. In ADNI, the estimated diseasetimeslines for 𝜖4 non-carriers and carriers were found to be significantly different from one another (𝑝 < 0.001).The estimate disease stage along the respective timelines distinguished AD subjects from controls with an AUCof 0.83 in both APOE 𝜖4 non-carriers and carriers. In the RS cohort, we obtained an AUC of 0.83 and 0.85 in 𝜖4non-carriers and carriers, respectively. Progression along the disease timelines as estimated by the rate of changeof disease stage showed a significant difference (𝑝 < 0.005) for subjects with pre-symptomatic AD as compared tothe general aging population in RS. It distinguished pre-symptomatic AD subjects with an AUC of 0.81 in APOE𝜖4 non-carriers and 0.88 in carriers, which was better than any individual volumetric biomarker, or its rate ofchange, could achieve. Our results suggest that co-init DEBM trained on case-controlled data is generalizable to apopulation-based cohort setting and that progression along the disease timelines is predictive of the developmentof AD in the general population. We expect that this approach can help to identify at-risk individuals from thegeneral population for targeted clinical trials as well as to provide biomarker based objective assessment in suchtrials.

Most existing algorithms for automatic 3D morphometry of human brain MRI scans are designed for data with near-isotropic voxels at approximately 1 mm resolution, and frequently have contrast constraints as well-typically requiring T1-weighted images (e.g., MP-RAGE scans). This limitation prevents the analysis of millions of MRI scans acquired with large inter-slice spacing in clinical settings every year. In turn, the inability to quantitatively analyze these scans hinders the adoption of quantitative neuro imaging in healthcare, and also precludes research studies that could attain huge sample sizes and hence greatly improve our understanding of the human brain. Recent advances in convolutional neural networks (CNNs) are producing outstanding results in super-resolution and contrast synthesis of MRI. However, these approaches are very sensitive to the specific combination of contrast, resolution and orientation of the input images, and thus do not generalize to diverse clinical acquisition protocols – even within sites. In this article, we present SynthSR, a method to train a CNN that receives one or more scans with spaced slices, acquired with different contrast, resolution and orientation, and produces an isotropic scan of canonical contrast (typically a 1 mm MP-RAGE). The presented method does not require any preprocessing, beyond rigid coregistration of the input scans. Crucially, SynthSR trains on synthetic input images generated from 3D segmentations, and can thus be used to train CNNs for any combination of contrasts, resolutions and orientations without high-resolution real images of the input contrasts. We test the images generated with SynthSR in an array of common downstream analyses, and show that they can be reliably used for subcortical segmentation and volumetry, image registration (e.g., for tensor-based morphometry), and, if some image quality requirements are met, even cortical thickness morphometry. The source code is publicly available at https://github.com/BBillot/SynthSR .

Alzheimer's disease (AD) is known as one of the major causes of dementia and is characterized by slow progression over several years, with no treatments or available medicines. In this regard, there have been efforts to identify the risk of developing AD in its earliest time. While many of the previous works considered cross-sectional analysis, more recent studies have focused on the diagnosis and prognosis of AD with longitudinal or time series data in a way of disease progression modeling. Under the same problem settings, in this work, we propose a novel computational framework that can predict the phenotypic measurements of MRI biomarkers and trajectories of clinical status along with cognitive scores at multiple future time points. However, in handling time series data, it generally faces many unexpected missing observations. In regard to such an unfavorable situation, we define a secondary problem of estimating those missing values and tackle it in a systematic way by taking account of temporal and multivariate relations inherent in time series data. Concretely, we propose a deep recurrent network that jointly tackles the four problems of (i) missing value imputation, (ii) phenotypic measurements forecasting, (iii) trajectory estimation of a cognitive score, and (iv) clinical status prediction of a subject based on his/her longitudinal imaging biomarkers. Notably, the learnable parameters of all the modules in our predictive models are trained in an end-to-end manner by taking the morphological features and cognitive scores as input, with our circumspectly defined loss function. In our experiments over The Alzheimers Disease Prediction Of Longitudinal Evolution (TADPOLE) challenge cohort, we measured performance for various metrics and compared our method to competing methods in the literature. Exhaustive analyses and ablation studies were also conducted to better confirm the effectiveness of our method.

Alzheimer’s disease (AD) is characterized by severe cognitive decline and pathological changes in the brain (brain atrophy, hyperphosphorylation of tau, and deposition of amyloid-beta protein). Females have greater neuropathology (AD biomarkers and brain atrophy rates) and cognitive decline than males, however these effects can depend on diagnosis (amnestic mild cognitive impairment (aMCI) or AD) and APOE genotype (presence of e4 alleles). Using the ADNI database (N = 630 females, N = 830 males), we analyzed the effect of sex, APOE genotype (non-carriers or carriers of APOEe4 alleles), and diagnosis (cognitively normal (CN), early aMCI (EMCI), late aMCI (LMCI), probable AD) on cognition (memory and executive function), hippocampal volume, and AD biomarkers (CSF levels of amyloid beta, tau, and ptau). Regardless of APOE genotype, memory scores were higher in CN, EMCI, and LMCI females compared to males but this sex difference was absent in probable AD, which may suggest a delay in the onset of cognitive decline or diagnosis and/or a faster trajectory of cognitive decline in females. We found that, regardless of diagnosis, CSF tau-pathology was disproportionately elevated in female carriers of APOEe4 alleles compared to males. In contrast, male carriers of APOEe4 alleles had reduced levels of CSF amyloid beta compared to females, irrespective of diagnosis. We also detected sex differences in hippocampal volume but the direction was dependent on the method of correction. Altogether results suggest that across diagnosis females show greater memory decline compared to males and APOE genotype affects AD neuropathology differently in males and females which may influence sex differences in incidence and progression of aMCI and AD.

To identify genetic variants influencing cortical atrophy in Alzheimer's disease (AD), we performed genome-wide association studies (GWAS) of mean cortical thicknesses in 17 AD-related brain. In this study, we used neuroimaging and genetic data of 919 participants from the Alzheimer's Disease Neuroimaging Initiative cohort, which include 268 cognitively normal controls, 488 mild cognitive impairment, 163 AD individuals. We performed GWAS with 3,041,429 single nucleotide polymorphisms (SNPs) for cortical thickness. The results of GWAS indicated that rs10109716 in ST18 (ST18 C2H2C-type zinc finger transcription factor) and rs661526 in NFIA (nuclear factor I A) genes are significantly associated with mean cortical thicknesses of the left inferior frontal gyrus and left parahippocampal gyrus, respectively. The rs661526 regulates the expression levels of NFIA in the substantia nigra and frontal cortex and rs10109716 regulates the expression levels of ST18 in the thalamus. These results suggest a crucial role of identified genes for cortical atrophy and could provide further insights into the genetic basis of AD.

In high-dimensional data analysis such as in genomics, repeated univariate regression for each variable is utilized to screen useful variables. However, signals jointly detectable with other variables may be overlooked. While the saturated model using all variables may not work in high-dimensional data, based on prior knowledge, group-wise analysis for a pre-defined group is often developed, but the power will be limited if the knowledge is insufficient. A flexible test procedure is thus proposed for conditional mean applicable to a variety of model sequences that bridge between low and high complexity models as in penalized regression. The test is based on the model that maximizes a generalization of the Yanai’s generalized coefficient of determination by exploiting the tendency for the dimensionality to be large under the null hypothesis. The test does not require complicated null distribution computation, thereby enabling large-scale testing application. Numerical studies demonstrated that the proposed test applied to the lasso and elastic net had a high power regardless of the simulation scenarios. Applied to a group-wise analysis in real genome-wide association study data from Alzheimer’s Disease Neuroimaging Initiative, the proposal gave a higher association signal than the existing methods.