Federated learning for medical image analysis: A survey.

Machine learning in medical imaging often faces a fundamental dilemma, namely, the small sample size problem. Many recent studies suggest using multi-domain data pooled from different acquisition sites/centers to improve statistical power. However, medical images from different sites cannot be easily shared to build large datasets for model training due to privacy protection reasons. As a promising solution, federated learning, which enables collaborative training of machine learning models based on data from different sites without cross-site data sharing, has attracted considerable attention recently. In this paper, we conduct a comprehensive survey of the recent development of federated learning methods in medical image analysis. We have systematically gathered research papers on federated learning and its applications in medical image analysis published between 2017 and 2023. Our search and compilation were conducted using databases from IEEE Xplore, ACM Digital Library, Science Direct, Springer Link, Web of Science, Google Scholar, and PubMed. In this survey, we first introduce the background of federated learning for dealing with privacy protection and collaborative learning issues. We then present a comprehensive review of recent advances in federated learning methods for medical image analysis. Specifically, existing methods are categorized based on three critical aspects of a federated learning system, including client end, server end, and communication techniques. In each category, we summarize the existing federated learning methods according to specific research problems in medical image analysis and also provide insights into the motivations of different approaches. In addition, we provide a review of existing benchmark medical imaging datasets and software platforms for current federated learning research. We also conduct an experimental study to empirically evaluate typical federated learning methods for medical image analysis. This survey can help to better understand the current research status, challenges, and potential research opportunities in this promising research field.

ATN cerebrospinal fluid biomarkers in dementia with Lewy bodies: Initial results from the United States Dementia with Lewy Bodies Consortium.

INTRODUCTION: The National Institute on Aging - Alzheimer's Association (NIA-AA) ATN research framework proposes to use biomarkers for amyloid (A), tau (T), and neurodegeneration (N) to stage individuals with AD pathological features and track changes longitudinally. The overall aim was to utilize this framework to characterize pre-mortem ATN status longitudinally in a clinically diagnosed cohort of dementia with Lewy bodies (DLB) and to correlate it with the post mortem diagnosis. METHODS: The cohort was subtyped by cerebrospinal fluid (CSF) ATN category. A subcohort had longitudinal data, and a subgroup was neuropathologically evaluated. RESULTS: We observed a significant difference in Aß42/40 after 12 months in the A+T- group. Post mortem neuropathologic analyses indicated that most of the p-Tau 181 positive (T+) cases also had a high Braak stage. DISCUSSION: This suggests that DLB patients who are A+ but T- may need to be monitored to determine whether they remain A+ or ever progress to T positivity. HIGHLIGHTS: Some A+T- DLB subjects transition from A+ to negative after 12-months. Clinically diagnosed DLB with LBP-AD (A+T+) maintain their positivity. Clinically diagnosed DLB with LBP-AD (A+T+) maintain their positivity. Monitoring of the A+T- sub-type of DLB may be necessary.

Plasma inflammation for predicting phenotypic conversion and clinical progression of autosomal dominant frontotemporal lobar degeneration.

BACKGROUND: Measuring systemic inflammatory markers may improve clinical prognosis and help identify targetable pathways for treatment in patients with autosomal dominant forms of frontotemporal lobar degeneration (FTLD). METHODS: We measured plasma concentrations of IL-6, TNFα and YKL-40 in pathogenic variant carriers (MAPT, C9orf72, GRN) and non-carrier family members enrolled in the ARTFL-LEFFTDS Longitudinal Frontotemporal Lobar Degeneration consortium. We evaluated associations between baseline plasma inflammation and rate of clinical and neuroimaging changes (linear mixed effects models with standardised (z) outcomes). We compared inflammation between asymptomatic carriers who remained clinically normal ('asymptomatic non-converters') and those who became symptomatic ('asymptomatic converters') using area under the curve analyses. Discrimination accuracy was compared with that of plasma neurofilament light chain (NfL). RESULTS: We studied 394 participants (non-carriers=143, C9orf72=117, GRN=62, MAPT=72). In MAPT, higher TNFα was associated with faster functional decline (B=0.12 (0.02, 0.22), p=0.02) and temporal lobe atrophy. In C9orf72, higher TNFα was associated with faster functional decline (B=0.09 (0.03, 0.16), p=0.006) and cognitive decline (B=-0.16 (-0.22, -0.10), p<0.001), while higher IL-6 was associated with faster functional decline (B=0.12 (0.03, 0.21), p=0.01). TNFα was higher in asymptomatic converters than non-converters (ß=0.29 (0.09, 0.48), p=0.004) and improved discriminability compared with plasma NfL alone (ΔR2=0.16, p=0.007; NfL: OR=1.4 (1.03, 1.9), p=0.03; TNFα: OR=7.7 (1.7, 31.7), p=0.007). CONCLUSIONS: Systemic proinflammatory protein measurement, particularly TNFα, may improve clinical prognosis in autosomal dominant FTLD pathogenic variant carriers who are not yet exhibiting severe impairment. Integrating TNFα with markers of neuronal dysfunction like NfL could optimise detection of impending symptom conversion in asymptomatic pathogenic variant carriers and may help personalise therapeutic approaches.

The effect of ß-amyloid positivity on cerebral metabolism in cognitively normal seniors.

INTRODUCTION: We evaluated the effect of cerebral amyloid-ß (Aß) deposition in cognitively normal (CN) seniors on regional metabolism of specific brain regions known to be affected by p-tau deposition. METHODS: Fluorodeoxyglucose positron emission tomography (FDG-PET), volumetric magnetic resonance imaging scans, and global amyloid standardized uptake value ratios (SUVr) were obtained for 210 CNs from the Alzheimer's Disease Neuroimaging Initiative-2 (ADNI2). Region of interest (ROI) extraction was used to obtain functional SUVr from six bilateral ROIs: amygdala (AM), entorhinal cortex (EC), hippocampus, lateral orbitofrontal, posterior cingulate (PC), and middle temporal gyrus. Every metabolic SUVr set was averaged and analyzed against the corresponding subject's amyloid SUVr. Correlation analyses were conducted on the full group and between APOE ε4-positive and APOE ε4-negative subgroups. RESULTS: The APOE ε4+ group exhibited significantly higher metabolism in the EC (r = 0.270, P = .038) and AM (r = 0.267, P = .041). When a significance of the difference test was conducted between the APOE ε4+ and APOE ε4-groups, these same regions remained significant: P = .012 and P = .016, respectively. By contrast, the APOE ε4 group displayed only the conventionally expected result of reduced regional metabolism in the PC (r = -0.161, P = .048), with higher Aß load. CONCLUSIONS: The effect of amyloid positivity on brain metabolism is regionally specific, and APOE ε4 status substantially modulates regional glucose uptake in these regions. The APOE ε4 allele may cause earlier emergence of clinical symptoms in AD via a mechanism that influences regional metabolic demand in specifically those regions where p-tau deposition is known to occur earliest.

Autonomic, Behavioral, and Subjective Pain Responses in Alzheimer's Disease.

OBJECTIVE: To compare autonomic, behavioral, and subjective pain responses of patients with Alzheimer's disease (AD) to those of healthy seniors (HS). As few studies have examined patients with severe Alzheimer's disease (sAD), we emphasized inclusion of these patients together with mild/moderate Alzheimer's disease (mAD) patients to characterize pain responses potentially affected by disease severity. DESIGN: A controlled cross-sectional study involving repeated measures behavioral pain testing. SETTING: An outpatient clinical setting and local nursing facilities. SUBJECTS: Community dwelling HS controls (N = 33) and individuals with chart-confirmed diagnoses of AD (N = 38, Diagnostic and Statistical Manual-IV criteria). METHODS: HS and AD groups were compared in their responses to repeated applications of five pressure intensities (1-5 kg) on the distal forearm. Autonomic responses (heart rate [HR]), pain behaviors (vocal, facial, and bodily as scored by the Pain Assessment in Advanced Dementia [PAINAD] scale), and subjective pain ratings (Faces Pain Scale-Revised) were measured. RESULTS: HR responses to pressure stimuli were differentially affected based on AD severity: sAD patients had generally decreased HR reactivity compared with other groups (P < 0.01). In contrast, pain behaviors were increased in AD regardless of severity (P < 0.001), compared with HS, for all but the lowest pressure intensity. Increased behaviors occurred in all measured domains of the PAINAD (P < 0.005). While sAD were unreliable subjective reporters, mAD patients (N = 17) rated low level pressures as more painful than HS (P < 0.01). CONCLUSION: These findings provide behavioral and subjective-report evidence of increased acute pain sensitivity in AD, which should be taken into consideration with respect to pain management across the spectrum of AD severity.

Examining Associations Between Smartphone Use and Clinical Severity in Frontotemporal Dementia: Proof-of-Concept Study.

BACKGROUND: Frontotemporal lobar degeneration (FTLD) is a leading cause of dementia in individuals aged <65 years. Several challenges to conducting in-person evaluations in FTLD illustrate an urgent need to develop remote, accessible, and low-burden assessment techniques. Studies of unobtrusive monitoring of at-home computer use in older adults with mild cognitive impairment show that declining function is reflected in reduced computer use; however, associations with smartphone use are unknown. OBJECTIVE: This study aims to characterize daily trajectories in smartphone battery use, a proxy for smartphone use, and examine relationships with clinical indicators of severity in FTLD. METHODS: Participants were 231 adults (mean age 52.5, SD 14.9 years; n=94, 40.7% men; n=223, 96.5% non-Hispanic White) enrolled in the Advancing Research and Treatment of Frontotemporal Lobar Degeneration (ARTFL study) and Longitudinal Evaluation of Familial Frontotemporal Dementia Subjects (LEFFTDS study) Longitudinal Frontotemporal Lobar Degeneration (ALLFTD) Mobile App study, including 49 (21.2%) with mild neurobehavioral changes and no functional impairment (ie, prodromal FTLD), 43 (18.6%) with neurobehavioral changes and functional impairment (ie, symptomatic FTLD), and 139 (60.2%) clinically normal adults, of whom 55 (39.6%) harbored heterozygous pathogenic or likely pathogenic variants in an autosomal dominant FTLD gene. Participants completed the Clinical Dementia Rating plus National Alzheimer's Coordinating Center Frontotemporal Lobar Degeneration Behavior and Language Domains (CDR+NACC FTLD) scale, a neuropsychological battery; the Neuropsychiatric Inventory; and brain magnetic resonance imaging. The ALLFTD Mobile App was installed on participants' smartphones for remote, passive, and continuous monitoring of smartphone use. Battery percentage was collected every 15 minutes over an average of 28 (SD 4.2; range 14-30) days. To determine whether temporal patterns of battery percentage varied as a function of disease severity, linear mixed effects models examined linear, quadratic, and cubic effects of the time of day and their interactions with each measure of disease severity on battery percentage. Models covaried for age, sex, smartphone type, and estimated smartphone age. RESULTS: The CDR+NACC FTLD global score interacted with time on battery percentage such that participants with prodromal or symptomatic FTLD demonstrated less change in battery percentage throughout the day (a proxy for less smartphone use) than clinically normal participants (P<.001 in both cases). Additional models showed that worse performance in all cognitive domains assessed (ie, executive functioning, memory, language, and visuospatial skills), more neuropsychiatric symptoms, and smaller brain volumes also associated with less battery use throughout the day (P<.001 in all cases). CONCLUSIONS: These findings support a proof of concept that passively collected data about smartphone use behaviors associate with clinical impairment in FTLD. This work underscores the need for future studies to develop and validate passive digital markers sensitive to longitudinal clinical decline across neurodegenerative diseases, with potential to enhance real-world monitoring of neurobehavioral change.

Reliability and Validity of Smartphone Cognitive Testing for Frontotemporal Lobar Degeneration.

Importance: Frontotemporal lobar degeneration (FTLD) is relatively rare, behavioral and motor symptoms increase travel burden, and standard neuropsychological tests are not sensitive to early-stage disease. Remote smartphone-based cognitive assessments could mitigate these barriers to trial recruitment and success, but no such tools are validated for FTLD. Objective: To evaluate the reliability and validity of smartphone-based cognitive measures for remote FTLD evaluations. Design, Setting, and Participants: In this cohort study conducted from January 10, 2019, to July 31, 2023, controls and participants with FTLD performed smartphone application (app)-based executive functioning tasks and an associative memory task 3 times over 2 weeks. Observational research participants were enrolled through 18 centers of a North American FTLD research consortium (ALLFTD) and were asked to complete the tests remotely using their own smartphones. Of 1163 eligible individuals (enrolled in parent studies), 360 were enrolled in the present study; 364 refused and 439 were excluded. Participants were divided into discovery (n = 258) and validation (n = 102) cohorts. Among 329 participants with data available on disease stage, 195 were asymptomatic or had preclinical FTLD (59.3%), 66 had prodromal FTLD (20.1%), and 68 had symptomatic FTLD (20.7%) with a range of clinical syndromes. Exposure: Participants completed standard in-clinic measures and remotely administered ALLFTD mobile app (app) smartphone tests. Main Outcomes and Measures: Internal consistency, test-retest reliability, association of smartphone tests with criterion standard clinical measures, and diagnostic accuracy. Results: In the 360 participants (mean [SD] age, 54.0 [15.4] years; 209 [58.1%] women), smartphone tests showed moderate-to-excellent reliability (intraclass correlation coefficients, 0.77-0.95). Validity was supported by association of smartphones tests with disease severity (r range, 0.38-0.59), criterion-standard neuropsychological tests (r range, 0.40-0.66), and brain volume (standardized ß range, 0.34-0.50). Smartphone tests accurately differentiated individuals with dementia from controls (area under the curve [AUC], 0.93 [95% CI, 0.90-0.96]) and were more sensitive to early symptoms (AUC, 0.82 [95% CI, 0.76-0.88]) than the Montreal Cognitive Assessment (AUC, 0.68 [95% CI, 0.59-0.78]) (z of comparison, -2.49 [95% CI, -0.19 to -0.02]; P = .01). Reliability and validity findings were highly similar in the discovery and validation cohorts. Preclinical participants who carried pathogenic variants performed significantly worse than noncarrier family controls on 3 app tasks (eg, 2-back ß = -0.49 [95% CI, -0.72 to -0.25]; P < .001) but not a composite of traditional neuropsychological measures (ß = -0.14 [95% CI, -0.42 to 0.14]; P = .32). Conclusions and Relevance: The findings of this cohort study suggest that smartphones could offer a feasible, reliable, valid, and scalable solution for remote evaluations of FTLD and may improve early detection. Smartphone assessments should be considered as a complementary approach to traditional in-person trial designs. Future research should validate these results in diverse populations and evaluate the utility of these tests for longitudinal monitoring.

Large-scale network analysis of the cerebrospinal fluid proteome identifies molecular signatures of frontotemporal lobar degeneration.

The pathophysiological mechanisms driving disease progression of frontotemporal lobar degeneration (FTLD) and corresponding biomarkers are not fully understood. We leveraged aptamer-based proteomics (> 4,000 proteins) to identify dysregulated communities of co-expressed cerebrospinal fluid proteins in 116 adults carrying autosomal dominant FTLD mutations (C9orf72, GRN, MAPT) compared to 39 noncarrier controls. Network analysis identified 31 protein co-expression modules. Proteomic signatures of genetic FTLD clinical severity included increased abundance of RNA splicing (particularly in C9orf72 and GRN) and extracellular matrix (particularly in MAPT) modules, as well as decreased abundance of synaptic/neuronal and autophagy modules. The generalizability of genetic FTLD proteomic signatures was tested and confirmed in independent cohorts of 1) sporadic progressive supranuclear palsy-Richardson syndrome and 2) frontotemporal dementia spectrum syndromes. Network-based proteomics hold promise for identifying replicable molecular pathways in adults living with FTLD. 'Hub' proteins driving co-expression of affected modules warrant further attention as candidate biomarkers and therapeutic targets.

Gene specific effects on brain volume and cognition of TMEM106B in frontotemporal lobar degeneration.

Background and Objectives: TMEM106B has been proposed as a modifier of disease risk in FTLD-TDP, particularly in GRN mutation carriers. Furthermore, TMEM106B has been investigated as a disease modifier in the context of healthy aging and across multiple neurodegenerative diseases. The objective of this study is to evaluate and compare the effect of TMEM106B on gray matter volume and cognition in each of the common genetic FTD groups and in sporadic FTD patients. Methods: Participants were enrolled through the ARTFL/LEFFTDS Longitudinal Frontotemporal Lobar Degeneration (ALLFTD) study, which includes symptomatic and presymptomatic individuals with a pathogenic mutation in C9orf72, GRN, MAPT, VCP, TBK1, TARDBP, symptomatic non-mutation carriers, and non-carrier family controls. All participants were genotyped for the TMEM106B rs1990622 SNP. Cross-sectionally, linear mixed-effects models were fitted to assess an association between TMEM106B and genetic group interaction with each outcome measure (gray matter volume and UDS3-EF for cognition), adjusting for education, age, sex and CDR®+NACC-FTLD sum of boxes. Subsequently, associations between TMEM106B and each outcome measure were investigated within the genetic group. For longitudinal modeling, linear mixed-effects models with time by TMEM106B predictor interactions were fitted. Results: The minor allele of TMEM106B rs1990622, linked to a decreased risk of FTD, associated with greater gray matter volume in GRN mutation carriers under the recessive dosage model. This was most pronounced in the thalamus in the left hemisphere, with a retained association when considering presymptomatic GRN mutation carriers only. The minor allele of TMEM106B rs1990622 also associated with greater cognitive scores among all C9orf72 mutation carriers and in presymptomatic C9orf72 mutation carriers, under the recessive dosage model. Discussion: We identified associations of TMEM106B with gray matter volume and cognition in the presence of GRN and C9orf72 mutations. This further supports TMEM106B as modifier of TDP-43 pathology. The association of TMEM106B with outcomes of interest in presymptomatic GRN and C9orf72 mutation carriers could additionally reflect TMEM106B's impact on divergent pathophysiological changes before the appearance of clinical symptoms.

Attention-Guided Autoencoder for Automated Progression Prediction of Subjective Cognitive Decline With Structural MRI.

Subjective cognitive decline (SCD) is the preclinical stage of Alzheimer's disease (AD) which happens even earlier than mild cognitive impairment (MCI). Progressive SCD will convert to MCI with the potential of further evolving to AD. Therefore, early identification of progressive SCD with neuroimaging techniques (e.g., structural MRI) is of great clinical value for early intervention of AD. However, existing MRI-based machine/deep learning methods usually suffer the small-sample-size problem and lack interpretability. To this end, we propose an interpretable autoencoder model with domain transfer learning (IADT) for progression prediction of SCD. Firstly, the proposed model can leverage MRIs from both the target domain (i.e., SCD) and auxiliary domains (e.g., AD and NC) for progressive SCD identification. Besides, it can automatically locate the disease-related brain regions of interest (defined in brain atlases) through an attention mechanism, which shows good interpretability. In addition, the IADT model is straightforward to train and test with only 5  âˆ¼ 10 seconds on CPUs and is suitable for medical tasks with small datasets. Extensive experiments on the publicly available ADNI dataset and a private CLAS dataset have demonstrated the effectiveness of the proposed method.

Hybrid Multimodality Fusion with Cross-Domain Knowledge Transfer to Forecast Progression Trajectories in Cognitive Decline.

Magnetic resonance imaging (MRI) and positron emission tomography (PET) are increasingly used to forecast progression trajectories of cognitive decline caused by preclinical and prodromal Alzheimer's disease (AD). Many existing studies have explored the potential of these two distinct modalities with diverse machine and deep learning approaches. But successfully fusing MRI and PET can be complex due to their unique characteristics and missing modalities. To this end, we develop a hybrid multimodality fusion (HMF) framework with cross-domain knowledge transfer for joint MRI and PET representation learning, feature fusion, and cognitive decline progression forecasting. Our HMF consists of three modules: 1) a module to impute missing PET images, 2) a module to extract multimodality features from MRI and PET images, and 3) a module to fuse the extracted multimodality features. To address the issue of small sample sizes, we employ a cross-domain knowledge transfer strategy from the ADNI dataset, which includes 795 subjects, to independent small-scale AD-related cohorts, in order to leverage the rich knowledge present within the ADNI. The proposed HMF is extensively evaluated in three AD-related studies with 272 subjects across multiple disease stages, such as subjective cognitive decline and mild cognitive impairment. Experimental results demonstrate the superiority of our method over several state-of-the-art approaches in forecasting progression trajectories of AD-related cognitive decline.

Disruption of limbic white matter pathways in mild cognitive impairment and Alzheimer's disease: a DTI/FDG-PET study.

BACKGROUND: Alzheimer's disease (AD) and mild cognitive impairment (MCI) affect the limbic system, causing medial temporal lobe (MTL) atrophy and posterior cingulate cortex (PCC) hypometabolism. Additionally, diffusion tensor imaging (DTI) studies have demonstrated that MCI and AD involve alterations in cerebral white matter (WM) integrity. OBJECTIVES: To test if (1) patients with MCI and AD exhibit decreases in the integrity of limbic WM pathways; (2) disconnection between PCC and MTL, manifested as disruption of the cingulum bundle, contributes to PCC hypometabolism during incipient AD. METHODS: We measured fractional anisotropy (FA) and volume of the fornix and cingulum using DTI in 23 individuals with MCI, 21 with mild-to-moderate AD, and 16 normal control (NC) subjects. We also measured PCC metabolism using (18)F-fluorodeoxyglucose positron emission tomography (FDG-PET) in AD and MCI patients. RESULTS: Fornix FA and volume were reduced in MCI and AD to a similar extent. Descending cingulum FA was reduced in AD while volume was reduced in MCI and even more so in AD. Both FA and volume of the fornix and descending cingulum reliably discriminated between NC and AD. Fornix FA and descending cingulum volume also reliably discriminated between NC and MCI. Only descending cingulum volume reliably discriminated between MCI and AD. In the combined MCI-AD cohort, PCC metabolism directly correlated with both FA and volume of the descending cingulum. CONCLUSIONS: Disruption of limbic WM pathways is evident during both MCI and AD. Disconnection of the PCC from MTL at the cingulum bundle contributes to PCC hypometabolism during incipient AD.

Comprehensive cross-sectional and longitudinal analyses of plasma neurofilament light across FTD spectrum disorders.

Frontotemporal dementia (FTD) therapy development is hamstrung by a lack of susceptibility, diagnostic, and prognostic biomarkers. Blood neurofilament light (NfL) shows promise as a biomarker, but studies have largely focused only on core FTD syndromes, often grouping patients with different diagnoses. To expedite the clinical translation of NfL, we avail ARTFL LEFFTDS Longitudinal Frontotemporal Lobar Degeneration (ALLFTD) study resources and conduct a comprehensive investigation of plasma NfL across FTD syndromes and in presymptomatic FTD mutation carriers. We find plasma NfL is elevated in all studied syndromes, including mild cases; increases in presymptomatic mutation carriers prior to phenoconversion; and associates with indicators of disease severity. By facilitating the identification of individuals at risk of phenoconversion, and the early diagnosis of FTD, plasma NfL can aid in participant selection for prevention or early treatment trials. Moreover, its prognostic utility would improve patient care, clinical trial efficiency, and treatment outcome estimations.

fMRI evidence that precision ophthalmic tints reduce cortical hyperactivation in migraine.

BACKGROUND: Certain patterns can induce perceptual illusions/distortions and visual discomfort in most people, headaches in patients with migraine, and seizures in patients with photosensitive epilepsy. Visual stimuli are common triggers for migraine attacks, possibly because of a hyperexcitability of the visual cortex shown in patients with migraine. Precision ophthalmic tints (POTs) are claimed to reduce perceptual distortions and visual discomfort and to prevent migraine headaches in some patients. We report an fMRI visual cortical activation study designed to investigate neurological mechanisms for the beneficial effects of POTs in migraine. METHODS: Eleven migraineurs and 11 age- and sex-matched non-headache controls participated in the study using non-stressful and stressful striped patterns viewed through gray, POT, and control coloured lenses. RESULTS: For all lenses, controls and migraineurs did not differ in their response to the non-stressful patterns. When the migraineurs wore gray lenses or control coloured lenses, the stressful pattern resulted in activation that was greater than in the controls. There was also an absence of the characteristic low-pass spatial frequency (SF) tuning in extrastriate visual areas. When POTs were worn, however, both cortical activation and SF tuning were normalized. Both when observing the stressful pattern and under more typical viewing conditions, the POTs reduced visual discomfort more than either of the other two lenses. CONCLUSION: The normalization of cortical activation and SF tuning in the migraineurs by POTs suggests a neurological basis for the therapeutic effect of these lenses in reducing visual cortical hyperactivation in migraine.

Subtyping of mild cognitive impairment using a deep learning model based on brain atrophy patterns.

Trajectories of cognitive decline vary considerably among individuals with mild cognitive impairment (MCI). To address this heterogeneity, subtyping approaches have been developed, with the objective of identifying more homogeneous subgroups. To date, subtyping of MCI has been based primarily on cognitive measures, often resulting in indistinct boundaries between subgroups and limited validity. Here, we introduce a subtyping method for MCI based solely upon brain atrophy. We train a deep learning model to differentiate between Alzheimer's disease (AD) and cognitively normal (CN) subjects based on whole-brain MRI features. We then deploy the trained model to classify MCI subjects based on whole-brain gray matter resemblance to AD-like or CN-like patterns. We subsequently validate the subtyping approach using cognitive, clinical, fluid biomarker, and molecular imaging data. Overall, the results suggest that atrophy patterns in MCI are sufficiently heterogeneous and can thus be used to subtype individuals into biologically and clinically meaningful subgroups.

Alzheimer's Disease Progressively Reduces Visual Functional Network Connectivity.

BACKGROUND: Postmortem studies of brains with Alzheimer's disease (AD) not only find amyloid-beta (Aß) and neurofibrillary tangles (NFT) in the visual cortex, but also reveal temporally sequential changes in AD pathology from higher-order association areas to lower-order areas and then primary visual area (V1) with disease progression. OBJECTIVE: This study investigated the effect of AD severity on visual functional network. METHODS: Eight severe AD (SAD) patients, 11 mild/moderate AD (MAD), and 26 healthy senior (HS) controls undertook a resting-state fMRI (rs-fMRI) and a task fMRI of viewing face photos. A resting-state visual functional connectivity (FC) network and a face-evoked visual-processing network were identified for each group. RESULTS: For the HS, the identified group-mean face-evoked visual-processing network in the ventral pathway started from V1 and ended within the fusiform gyrus. In contrast, the resting-state visual FC network was mainly confined within the visual cortex. AD disrupted these two functional networks in a similar severity dependent manner: the more severe the cognitive impairment, the greater reduction in network connectivity. For the face-evoked visual-processing network, MAD disrupted and reduced activation mainly in the higher-order visual association areas, with SAD further disrupting and reducing activation in the lower-order areas. CONCLUSION: These findings provide a functional corollary to the canonical view of the temporally sequential advancement of AD pathology through visual cortical areas. The association of the disruption of functional networks, especially the face-evoked visual-processing network, with AD severity suggests a potential predictor or biomarker of AD progression.

Fluid and Tissue Biomarkers of Lewy Body Dementia: Report of an LBDA Symposium.

The Lewy Body Dementia Association (LBDA) held a virtual event, the LBDA Biofluid/Tissue Biomarker Symposium, on January 25, 2021, to present advances in biomarkers for Lewy body dementia (LBD), which includes dementia with Lewy bodies (DLBs) and Parkinson's disease dementia (PDD). The meeting featured eight internationally known scientists from Europe and the United States and attracted over 200 scientists and physicians from academic centers, the National Institutes of Health, and the pharmaceutical industry. Methods for confirming and quantifying the presence of Lewy body and Alzheimer's pathology and novel biomarkers were discussed.

Alzheimer's Disease Progressively Alters the Face-Evoked Visual-Processing Network.

BACKGROUND: Postmortem studies of Alzheimer's disease (AD) brains not only find amyloid-ß (Aß) and neurofibrillary tangles (NFT) in the primary and associative visual cortical areas, but also reveal a temporally successive sequence of AD pathology beginning in higher-order visual association areas, followed by involvement of lower-order visual processing regions with disease progression, and extending to primary visual cortex in late-stage disease. These findings suggest that neuronal loss associated with Aß and NFT aggregation in these areas may alter not only the local neuronal activation but also visual neural network activity. OBJECTIVE: Applying a novel method to identify the visual functional network and investigate the association of the network changes with disease progression. METHODS: To investigate the effect of AD on the face-evoked visual-processing network, 8 severe AD (SAD) patients, 11 mild/moderate AD (MAD), and 26 healthy senior (HS) controls undertook a task-fMRI study of viewing face photos. RESULTS: For the HS, the identified group-mean visual-processing network in the ventral pathway started from V1 and ended within the fusiform gyrus. In contrast, this network was disrupted and reduced in the AD patients in a disease-severity dependent manner: for the MAD patients, the network was disrupted and reduced mainly in the higher-order visual association areas; for the SAD patients, the network was nearly absent in the higher-order association areas, and disrupted and reduced in the lower-order areas. CONCLUSION: This finding is consistent with the current canonical view of the temporally successive sequence of AD pathology through visual cortical areas.

Heterogeneity in mild cognitive impairment: differences in neuropsychological profile and associated white matter lesion pathology.

This study examined whether distinct neuropsychological profiles could be delineated in a sample with Mild Cognitive Impairment (MCI) and whether white matter lesion (WML) burden contributed to MCI group differences. A heterogeneous, clinical sample of 70 older adults diagnosed with MCI was assessed using cognitive scores, and WML was quantified using a semi-automated, volumetric approach on T2-weighted fluid-attenuated inversion recovery (FLAIR) images. Using cluster and discriminant analyses, three distinct groups (Memory/Language, Executive/Processing Speed, and Pure Memory) were empirically derived based on cognitive scores. Results also showed a dose dependent relationship of WML burden to MCI subgroup, with the Executive/Processing Speed subgroup demonstrating significantly higher levels of WML pathology when compared to the other subgroups. In addition, there was a dissociation of lesion type by the two most impaired subgroups (Memory/Language and Executive/Processing Speed) such that the Memory/Language subgroup showed higher periventricular lesion (PVL) and lower deep white matter lesion (DWML) volumes, whereas the Executive/Processing Speed demonstrated higher DWML and lower PVL volumes. Results demonstrate that distinct MCI subgroups can be empirically derived and reliably differentiated from a heterogeneous MCI sample, and that these profiles differ according to WML burden. Overall, findings suggest different underlying pathologies within MCI and contribute to our understanding of MCI subtypes.