Explore the vast range of titles available.

Dementia is characterized by a decline in memory and thinking that is significant enough to impair function in activities of daily living. Patients seen in dementia specialty clinics are highly heterogenous with a variety of different symptoms that progress at different rates. Recent research has focused on finding data-driven subtypes for revealing new insights into dementia’s underlying heterogeneity, rather than assuming that the cohort is homogenous. However, current studies on dementia subtyping have the following limitations: (i) focusing on AD-related dementia only and not examining heterogeneity within dementia as a whole, (ii) using only cross-sectional baseline visit information for clustering and (iii) predominantly relying on expensive imaging biomarkers as features for clustering. In this study, we seek to overcome such limitations, using a data-driven unsupervised clustering algorithm named SillyPutty, in combination with hierarchical clustering on cognitive assessment scores to estimate subtypes within a real-world clinical dementia cohort. We use a longitudinal patient data set for our clustering analysis, instead of relying only on baseline visits, allowing us to explore the ongoing temporal relationship between subtypes and disease progression over time. Results showed that subtypes with very mild or mild dementia were more heterogenous in their cognitive profiles and risk of disease progression.

Differences in amyloid positron emission tomography (PET) radiotracer pharmacokinetics and binding properties lead to discrepancies in amyloid‐β uptake estimates. Harmonization of tracer‐specific biases is crucial for optimal performance of downstream tasks. Here, we investigated the efficacy of ComBat, a data‐driven harmonization model, for reducing tracer‐specific biases in regional amyloid PET measurements from [18F]‐florbetapir (FBP) and [11C]‐Pittsburgh compound‐B (PiB). One hundred thirteen head‐to‐head FBP‐PiB scan pairs, scanned from the same subject within 90 days, were selected from the Open Access Series of Imaging Studies 3 (OASIS‐3) dataset. The Centiloid scale, ComBat with no covariates, ComBat with biological covariates, and GAM‐ComBat with biological covariates were used to harmonize both global and regional amyloid standardized uptake value ratios (SUVR). Variants of ComBat, including longitudinal ComBat and PEACE, were also tested. Intraclass correlation coefficient (ICC) and mean absolute error (MAE) were computed to measure the absolute agreement between tracers. Additionally, longitudinal amyloid SUVRs from an anti‐amyloid drug trial were simulated using linear mixed effects modeling. Differences in rates‐of‐change between simulated treatment and placebo groups were tested, and change in statistical power/Type‐I error after harmonization was quantified. In the head‐to‐head tracer comparison, ComBat with no covariates was the best at increasing ICC and decreasing MAE of both global summary and regional amyloid PET SUVRs between scan pairs of the same group of subjects. In the clinical trial simulation, harmonization with both Centiloid and ComBat increased statistical power of detecting true rate‐of‐change differences between groups and decreased false discovery rate in the absence of a treatment effect. The greatest benefit of harmonization was observed when groups exhibited differing FBP‐to‐PiB proportions. ComBat outperformed the Centiloid scale in harmonizing both global and regional amyloid estimates. Additionally, ComBat improved the detection of rate‐of‐change differences between clinical trial groups. Our findings suggest that ComBat is a viable alternative to Centiloid for harmonizing regional amyloid PET analyses. Using paired amyloid PET scans of the same subject using two different radiotracers, we showed that ComBat, a data‐driven harmonization technique, improves the absolute agreement of regional PET measurements across tracers. We also conducted simulations to demonstrate that ComBat may help increase statistical power in anti‐amyloid drugs clinical trials.

Alzheimer’s Disease (AD) is the leading cause of dementia, characterized by cognitive and functional impairments that disrupt daily activities. Different clinical modalities such as neuroimaging biomarkers, cognitive assessments, fluid biomarkers and genetic data provide unique and complementary information, contributing to a more comprehensive understanding of disease progression and heterogeneity in disease characteristics. With recent advancements in computational capabilities, particularly in deep learning, multimodal representation learning frameworks aim to integrate diverse clinical modalities into a cohesive framework, capturing the most significant patterns within each modality. Existing data-driven multimodal representation learning frameworks in AD research have two major limitations. First, AD progresses gradually from early preclinical stages to severe impairment, requiring dynamic, longitudinal assessments for timely interventions, rather than single-endpoint predictions. Additionally, its significant heterogeneity in symptoms, progression, and pathology is often overlooked by traditional ML/DL methods, which rely on group averages. Precision medicine demands a shift toward characterizing disease abnormalities at the individual level.Towards this end, three research aims were pursued. (1) First, we developed HiMAL, a novel multimodal Hierarchical Multi-task Auxiliary Learning framework that predicts cognitive composite scores as auxiliary tasks to estimate the longitudinal risk of progression from Mild Cognitive Impairment (MCI) to Alzheimer’s Disease (AD). HiMAL also provides interpretable, longitudinal explanations of disease progression to support clinical decision-making (2) Next, we designed data-driven unsupervised machine learning frameworks to explore inter-individual heterogeneity within AD cohorts. This includes an unsupervised clustering framework leveraging cognitive assessments and neuroimaging biomarkers to identify subtypes within heterogeneous dementia populations. Additionally, we developed a deep learning-based normative modeling framework to examine AD heterogeneity using multimodal neuroimaging biomarkers of neuropathology and neurodegeneration.(3) Finally, building on this, we created a multimodal introspective variational autoencoder to enhance normative modeling and improve the detection of individual-level disease abnormalities.The three aims in this dissertation collectively advances the state of multimodal representation learning by developing computational frameworks to address two key challenges of heterogeneity and progression in AD. While Aim 1 establishes a multimodal longitudinal framework to predict cognitive decline, setting the stage for personalized disease monitoring, Aims 2 and 3 extends this foundation by investigating inter-individual variability using unsupervised frameworks, delving deeper into the heterogeneous nature of AD. Together, these aims form a cohesive narrative, advancing our understanding of AD and proposing tools for precision diagnostics.

Background Differences in amyloid PET radiotracer pharmacokinetics and binding properties lead to discrepancies in amyloid uptake measurements, which may adversely affect the statistical power of clinical trials that utilize multiple tracers to track brain amyloid deposition. To address this, Centiloid was developed for standardizing global amyloid SUVRs across tracers to a common scale. Alternatively, ComBat is a technique for harmonizing batch effects while preserving variations from biologically‐relevant covariates. Unlike Centiloid, ComBat is entirely data‐driven, does not require a calibration process, and can be applied to regional SUVRs. However, it has not been validated for amyloid PET. Here, we evaluate whether ComBat improves detection of treatment effects in a simulated clinical trial. Method 365 amyloid‐positive subjects were identified from the OASIS‐3 dataset, from which 322 Pittsburgh compound B (PiB) and 260 18F‐Florbetapir (FBP) PET scans were selected. SUVRs of the global cortical region and 82 FreeSurfer regions were computed from each scan using the PET Unified Pipeline. Linear mixed effects (LME) models were fitted on each tracer separately and used to simulate longitudinal SUVRs of placebo and treatment groups in a hypothetical clinical trial (Figure 1). A negative rate‐of‐change was introduced to the treatment group to mimic a treatment effect. Additionally, tracer mixing proportions were varied within each group. Centiloid and ComBat were applied to simulated data to harmonize across tracers. A separate LME model was fitted to test for significant differences in SUVR rate‐of‐change between groups. Power was estimated as the proportion of significant findings across 1000 simulations. Result After harmonization with either Centiloid or ComBat with no covariates, an increase in power was observed in the presence of a treatment effect, and a decrease in Type‐I error was observed for no treatment effect (Figure 2). These changes were most prominent in cases where groups exhibited differing tracer mixing proportions. Similar patterns were observed for regional SUVRs (Figure 3). Conclusion We demonstrated that tracer harmonization is important for improving power in the presence of a treatment effect and reducing Type‐I error in its absence. ComBat performs comparably to Centiloid in harmonizing amyloid radiotracers in the context of a clinical trial.

Background Differences in amyloid PET radiotracer pharmacokinetics and binding properties lead to discrepancies in amyloid uptake measurements, which may adversely affect the statistical power of clinical trials that utilize multiple tracers to track brain amyloid deposition. To address this, Centiloid was developed for standardizing global amyloid SUVRs across tracers to a common scale. Alternatively, ComBat is a technique for harmonizing batch effects while preserving variations from biologically‐relevant covariates. Unlike Centiloid, ComBat is entirely data‐driven, does not require a calibration process, and can be applied to regional SUVRs. However, it has not been validated for amyloid PET. Here, we evaluate whether ComBat improves detection of treatment effects in a simulated clinical trial. Method 365 amyloid‐positive subjects were identified from the OASIS‐3 dataset, from which 322 Pittsburgh compound B (PiB) and 260 18F‐Florbetapir (FBP) PET scans were selected. SUVRs of the global cortical region and 82 FreeSurfer regions were computed from each scan using the PET Unified Pipeline. Linear mixed effects (LME) models were fitted on each tracer separately and used to simulate longitudinal SUVRs of placebo and treatment groups in a hypothetical clinical trial (Figure 1). A negative rate‐of‐change was introduced to the treatment group to mimic a treatment effect. Additionally, tracer mixing proportions were varied within each group. Centiloid and ComBat were applied to simulated data to harmonize across tracers. A separate LME model was fitted to test for significant differences in SUVR rate‐of‐change between groups. Power was estimated as the proportion of significant findings across 1000 simulations. Result After harmonization with either Centiloid or ComBat with no covariates, an increase in power was observed in the presence of a treatment effect, and a decrease in Type‐I error was observed for no treatment effect (Figure 2). These changes were most prominent in cases where groups exhibited differing tracer mixing proportions. Similar patterns were observed for regional SUVRs (Figure 3). Conclusion We demonstrated that tracer harmonization is important for improving power in the presence of a treatment effect and reducing Type‐I error in its absence. ComBat performs comparably to Centiloid in harmonizing amyloid radiotracers in the context of a clinical trial.

INTRODUCTION Previous studies have applied normative modeling on a single neuroimaging modality to investigate Alzheimer disease (AD) heterogeneity. We employed a deep learning‐based multimodal normative framework to analyze individual‐level variation across ATN (amyloid‐tau‐neurodegeneration) imaging biomarkers. METHODS We selected cross‐sectional discovery (n = 665) and replication cohorts (n = 430) with available T1‐weighted magnetic resonance imaging (MRI), amyloid, and tau positron emission tomography (PET). Normative modeling estimated individual‐level abnormal deviations in amyloid‐positive individuals compared to amyloid‐negative controls. Regional abnormality patterns were mapped at different clinical group levels to assess intra‐group heterogeneity. An individual‐level disease severity index (DSI) was calculated using both the spatial extent and magnitude of abnormal deviations across ATN. RESULTS Greater intra‐group heterogeneity in ATN abnormality patterns was observed in more severe clinical stages of AD. Higher DSI was associated with worse cognitive function and increased risk of disease progression. DISCUSSION Subject‐specific abnormality maps across ATN reveal the heterogeneous impact of AD on the brain. Highlights Normative modeling examined AD heterogeneity across multimodal imaging biomarkers. Heterogeneity in spatial patterns of gray matter atrophy, amyloid, and tau burden. Higher within‐group heterogeneity for AD patients at advanced dementia stages. Patient‐specific metric summarized extent of neurodegeneration and neuropathology. Metric is a marker of poor brain health and can monitor risk of disease progression.

Objective We aimed to develop and validate a novel multimodal framework Hierarchical Multi-task Auxiliary Learning (HiMAL) framework, for predicting cognitive composite functions as auxiliary tasks that estimate the longitudinal risk of transition from Mild Cognitive Impairment (MCI) to Alzheimer’s Disease (AD). Materials and Methods HiMAL utilized multimodal longitudinal visit data including imaging features, cognitive assessment scores, and clinical variables from MCI patients in the Alzheimer’s Disease Neuroimaging Initiative dataset, to predict at each visit if an MCI patient will progress to AD within the next 6 months. Performance of HiMAL was compared with state-of-the-art single-task and multitask baselines using area under the receiver operator curve (AUROC) and precision recall curve (AUPRC) metrics. An ablation study was performed to assess the impact of each input modality on model performance. Additionally, longitudinal explanations regarding risk of disease progression were provided to interpret the predicted cognitive decline. Results Out of 634 MCI patients (mean [IQR] age: 72.8 [67-78], 60% male), 209 (32%) progressed to AD. HiMAL showed better prediction performance compared to all state-of-the-art longitudinal single-modality singe-task baselines (AUROC = 0.923 [0.915-0.937]; AUPRC = 0.623 [0.605-0.644]; all P < .05). Ablation analysis highlighted that imaging and cognition scores with maximum contribution towards prediction of disease progression. Discussion Clinically informative model explanations anticipate cognitive decline 6 months in advance, aiding clinicians in future disease progression assessment. HiMAL relies on routinely collected electronic health records (EHR) variables for proximal (6 months) prediction of AD onset, indicating its translational potential for point-of-care monitoring and managing of high-risk patients. Lay Summary Alzheimer’s Disease (AD), neurodegenerative disorder, results in memory loss and cognitive decline, affecting millions of individuals globally. For patients with mild cognitive impairment (MCI), a prodromal phrase of AD, it is important to monitor their risk of progressing to AD over time. Using longitudinal multimodal data on imaging, cognition and clinical data from a publicly available AD dataset, we developed and validated HiMAL a novel multimodal Hierarchical Multi-task Auxiliary Learning framework to estimate at every visit timepoint if a MCI patient will progress to AD within the next 6 months. HiMAL predicted forecasted neuropsychological composite function scores as auxiliary tasks at every timepoint and used the predicted forecasted scores to the predict the main task of predicting progression. The hierarchical approach allows HiMAL to effectively use task dependencies, resulting in more informative features and better performance than state-of-the-art multimodal and multi-task methods. Additionally, HiMAL provided longitudinal explanations that can inform clinicians 6 months in advance the potential cognitive function decline that can lead to progression to AD in future. Built on routinely collected EHR data, HiMAL's flexibility in the selection of input modalities and auxiliary tasks suggest that it can be applied to various AD datasets and other clinical problems.