Unsupervised multimodal modeling of cognitive and brain health trajectories for early dementia prediction.

Predicting the course of neurodegenerative disorders early has potential to greatly improve clinical management and patient outcomes. A key challenge for early prediction in real-world clinical settings is the lack of labeled data (i.e., clinical diagnosis). In contrast to supervised classification approaches that require labeled data, we propose an unsupervised multimodal trajectory modeling (MTM) approach based on a mixture of state space models that captures changes in longitudinal data (i.e., trajectories) and stratifies individuals without using clinical diagnosis for model training. MTM learns the relationship between states comprising expensive, invasive biomarkers (ß-amyloid, grey matter density) and readily obtainable cognitive observations. MTM training on trajectories stratifies individuals into clinically meaningful clusters more reliably than MTM training on baseline data alone and is robust to missing data (i.e., cognitive data alone or single assessments). Extracting an individualized cognitive health index (i.e., MTM-derived cluster membership index) allows us to predict progression to AD more precisely than standard clinical assessments (i.e., cognitive tests or MRI scans alone). Importantly, MTM generalizes successfully from research cohort to real-world clinical data from memory clinic patients with missing data, enhancing the clinical utility of our approach. Thus, our multimodal trajectory modeling approach provides a cost-effective and non-invasive tool for early dementia prediction without labeled data (i.e., clinical diagnosis) with strong potential for translation to clinical practice.

Using Repeated Measurements to Predict Cardiovascular Risk in Patients With Type 2 Diabetes Mellitus.

The QRISK cardiovascular disease (CVD) risk assessment model is not currently optimized for patients with type 2 diabetes mellitus (T2DM). We aim to identify if the abundantly available repeatedly measured data for patients with T2D improves the predictive capability of QRISK to support the decision-making process regarding CVD prevention in patients with T2DM. We identified patients with T2DM aged 25 to 85, not on statin treatment and without pre-existing CVD from the IQVIA Medical Research Data United Kingdom primary care database and then followed them up until the first diagnosis of CVD, ischemic heart disease, or stroke/transient ischemic attack. We included traditional, nontraditional risk factors and relevant treatments for our analysis. We then undertook a Cox's hazards model accounting for time-dependent covariates to estimate the hazard rates for each risk factor and calculated a 10-year risk score. Models were developed for males and females separately. We tested the performance of our models using validation data and calculated discrimination and calibration statistics. The study included 198,835 (180,143 male with 11,976 outcomes and 90,466 female with 8,258 outcomes) patients. The 10-year predicted survival probabilities for females was 0.87 (0.87 to 0.87), whereas the observed survival estimates from the Kaplan-Meier curve for all female models was 0.87 (0.86 to 0.87). The predicted and observed survival estimates for males were 0.84 (0.84 to 0.84) and 0.84 (0.83 to 0.84) respectively. The Harrell's C-index of all female models and all male models were 0.71 and 0.69 respectively. We found that including time-varying repeated measures, only mildly improved CVD risk prediction for T2DM patients in comparison to the current practice standard. We advocate for further research using time-varying data to identify if the involvement of further covariates may improve the accuracy of currently accepted prediction models.

Emerging opportunities and challenges for the future of reservoir computing.

Reservoir computing originates in the early 2000s, the core idea being to utilize dynamical systems as reservoirs (nonlinear generalizations of standard bases) to adaptively learn spatiotemporal features and hidden patterns in complex time series. Shown to have the potential of achieving higher-precision prediction in chaotic systems, those pioneering works led to a great amount of interest and follow-ups in the community of nonlinear dynamics and complex systems. To unlock the full capabilities of reservoir computing towards a fast, lightweight, and significantly more interpretable learning framework for temporal dynamical systems, substantially more research is needed. This Perspective intends to elucidate the parallel progress of mathematical theory, algorithm design and experimental realizations of reservoir computing, and identify emerging opportunities as well as existing challenges for large-scale industrial adoption of reservoir computing, together with a few ideas and viewpoints on how some of those challenges might be resolved with joint efforts by academic and industrial researchers across multiple disciplines.

A two-step inverse solution for a single dipole cardiac source.

Introduction: The inverse problem of electrocardiography noninvasively localizes the origin of undesired cardiac activity, such as a premature ventricular contraction (PVC), from potential recordings from multiple torso electrodes. However, the optimal number and placement of electrodes for an accurate solution of the inverse problem remain undetermined. This study presents a two-step inverse solution for a single dipole cardiac source, which investigates the significance of the torso electrodes on a patient-specific level. Furthermore, the impact of the significant electrodes on the accuracy of the inverse solution is studied. Methods: Body surface potential recordings from 128 electrodes of 13 patients with PVCs and their corresponding homogeneous and inhomogeneous torso models were used. The inverse problem using a single dipole was solved in two steps: First, using information from all electrodes, and second, using a subset of electrodes sorted in descending order according to their significance estimated by a greedy algorithm. The significance of electrodes was computed for three criteria derived from the singular values of the transfer matrix that correspond to the inversely estimated origin of the PVC computed in the first step. The localization error (LE) was computed as the Euclidean distance between the ground truth and the inversely estimated origin of the PVC. The LE obtained using the 32 and 64 most significant electrodes was compared to the LE obtained when all 128 electrodes were used for the inverse solution. Results: The average LE calculated for both torso models and using all 128 electrodes was 28.8 ± 11.9 mm. For the three tested criteria, the average LEs were 32.6 ± 19.9 mm, 29.6 ± 14.7 mm, and 28.8 ± 14.5 mm when 32 electrodes were used. When 64 electrodes were used, the average LEs were 30.1 ± 16.8 mm, 29.4 ± 12.0 mm, and 29.5 ± 12.6 mm. Conclusion: The study found inter-patient variability in the significance of torso electrodes and demonstrated that an accurate localization by the inverse solution with a single dipole could be achieved using a carefully selected reduced number of electrodes.

Flexible structure learning under uncertainty.

Experience is known to facilitate our ability to interpret sequences of events and make predictions about the future by extracting temporal regularities in our environments. Here, we ask whether uncertainty in dynamic environments affects our ability to learn predictive structures. We exposed participants to sequences of symbols determined by first-order Markov models and asked them to indicate which symbol they expected to follow each sequence. We introduced uncertainty in this prediction task by manipulating the: (a) probability of symbol co-occurrence, (b) stimulus presentation rate. Further, we manipulated feedback, as it is known to play a key role in resolving uncertainty. Our results demonstrate that increasing the similarity in the probabilities of symbol co-occurrence impaired performance on the prediction task. In contrast, increasing uncertainty in stimulus presentation rate by introducing temporal jitter resulted in participants adopting a strategy closer to probability maximization than matching and improving in the prediction tasks. Next, we show that feedback plays a key role in learning predictive statistics. Trial-by-trial feedback yielded stronger improvement than block feedback or no feedback; that is, participants adopted a strategy closer to probability maximization and showed stronger improvement when trained with trial-by-trial feedback. Further, correlating individual strategy with learning performance showed better performance in structure learning for observers who adopted a strategy closer to maximization. Our results indicate that executive cognitive functions (i.e., selective attention) may account for this individual variability in strategy and structure learning ability. Taken together, our results provide evidence for flexible structure learning; individuals adapt their decision strategy closer to probability maximization, reducing uncertainty in temporal sequences and improving their ability to learn predictive statistics in variable environments.

Influence and influenceability: global directionality in directed complex networks.

Knowing which nodes are influential in a complex network and whether the network can be influenced by a small subset of nodes is a key part of network analysis. However, many traditional measures of importance focus on node level information without considering the global network architecture. We use the method of trophic analysis to study directed networks and show that both 'influence' and 'influenceability' in directed networks depend on the hierarchical structure and the global directionality, as measured by the trophic levels and trophic coherence, respectively. We show that in directed networks trophic hierarchy can explain: the nodes that can reach the most others; where the eigenvector centrality localizes; which nodes shape the behaviour in opinion or oscillator dynamics; and which strategies will be successful in generalized rock-paper-scissors games. We show, moreover, that these phenomena are mediated by the global directionality. We also highlight other structural properties of real networks related to influenceability, such as the pseudospectra, which depend on trophic coherence. These results apply to any directed network and the principles highlighted-that node hierarchy is essential for understanding network influence, mediated by global directionality-are applicable to many real-world dynamics.

Identification of Clinically Relevant HIV Vif Protein Motif Mutations through Machine Learning and Undersampling.

Human Immunodeficiency virus (HIV) and its clinical entity, the Acquired Immunodeficiency Syndrome (AIDS) continue to represent an important health burden worldwide. Although great advances have been made towards determining the way viral genetic diversity affects clinical outcome, genetic association studies have been hindered by the complexity of their interactions with the human host. This study provides an innovative approach for the identification and analysis of epidemiological associations between HIV Viral Infectivity Factor (Vif) protein mutations and four clinical endpoints (Viral load and CD4 T cell numbers at time of both clinical debut and on historical follow-up of patients. Furthermore, this study highlights an alternative approach to the analysis of imbalanced datasets, where patients without specific mutations outnumber those with mutations. Imbalanced datasets are still a challenge hindering the development of classification algorithms through machine learning. This research deals with Decision Trees, Naïve Bayes (NB), Support Vector Machines (SVMs), and Artificial Neural Networks (ANNs). This paper proposes a new methodology considering an undersampling approach to deal with imbalanced datasets and introduces two novel and differing approaches (MAREV-1 and MAREV-2). As theses approaches do not involve human pre-determined and hypothesis-driven combinations of motifs having functional or clinical relevance, they provide a unique opportunity to discover novel complex motif combinations of interest. Moreover, the motif combinations found can be analyzed through traditional statistical approaches avoiding statistical corrections for multiple tests.

Strong connectivity in real directed networks.

In many real, directed networks, the strongly connected component of nodes which are mutually reachable is very small. This does not fit with current theory, based on random graphs, according to which strong connectivity depends on mean degree and degree-degree correlations. And it has important implications for other properties of real networks and the dynamical behavior of many complex systems. We find that strong connectivity depends crucially on the extent to which the network has an overall direction or hierarchical ordering-a property measured by trophic coherence. Using percolation theory, we find the critical point separating weakly and strongly connected regimes and confirm our results on many real-world networks, including ecological, neural, trade, and social networks. We show that the connectivity structure can be disrupted with minimal effort by a targeted attack on edges which run counter to the overall direction. This means that many dynamical processes on networks can depend significantly on a small fraction of edges.

Machine learning classification of conduct disorder with high versus low levels of callous-unemotional traits based on facial emotion recognition abilities.

Conduct disorder (CD) with high levels of callous-unemotional traits (CD/HCU) has been theoretically linked to specific difficulties with fear and sadness recognition, in contrast to CD with low levels of callous-unemotional traits (CD/LCU). However, experimental evidence for this distinction is mixed, and it is unclear whether these difficulties are a reliable marker of CD/HCU compared to CD/LCU. In a large sample (N = 1263, 9-18 years), we combined univariate analyses and machine learning classifiers to investigate whether CD/HCU is associated with disproportionate difficulties with fear and sadness recognition over other emotions, and whether such difficulties are a reliable individual-level marker of CD/HCU. We observed similar emotion recognition abilities in CD/HCU and CD/LCU. The CD/HCU group underperformed relative to typically developing (TD) youths, but difficulties were not specific to fear or sadness. Classifiers did not distinguish between youths with CD/HCU versus CD/LCU (52% accuracy), although youths with CD/HCU and CD/LCU were reliably distinguished from TD youths (64% and 60%, respectively). In the subset of classifiers that performed well for youths with CD/HCU, fear and sadness were the most relevant emotions for distinguishing them from youths with CD/LCU and TD youths, respectively. We conclude that non-specific emotion recognition difficulties are common in CD/HCU, but are not reliable individual-level markers of CD/HCU versus CD/LCU. These findings highlight that a reduced ability to recognise facial expressions of distress should not be assumed to be a core feature of CD/HCU.

Generalized Learning Vector Quantization With Log-Euclidean Metric Learning on Symmetric Positive-Definite Manifold.

In many classification scenarios, the data to be analyzed can be naturally represented as points living on the curved Riemannian manifold of symmetric positive-definite (SPD) matrices. Due to its non-Euclidean geometry, usual Euclidean learning algorithms may deliver poor performance on such data. We propose a principled reformulation of the successful Euclidean generalized learning vector quantization (GLVQ) methodology to deal with such data, accounting for the nonlinear Riemannian geometry of the manifold through log-Euclidean metric (LEM). We first generalize GLVQ to the manifold of SPD matrices by exploiting the LEM-induced geodesic distance (GLVQ-LEM). We then extend GLVQ-LEM with metric learning. In particular, we study both 1) a more straightforward implementation of the metric learning idea by adapting metric in the space of vectorized log-transformed SPD matrices and 2) the full formulation of metric learning without matrix vectorization, thus preserving the second-order tensor structure. To obtain the distance metric in the full LEM learning (LEML) approaches, two algorithms are proposed. One method is to restrict the distance metric to be full rank, treating the distance metric tensor as an SPD matrix, and readily use the LEM framework (GLVQ-LEML-LEM). The other method is to cast no such restriction, treating the distance metric tensor as a fixed rank positive semidefinite matrix living on a quotient manifold with total space equipped with flat geometry (GLVQ-LEML-FM). Experiments on multiple datasets of different natures demonstrate the good performance of the proposed methods.

Network hierarchy and pattern recovery in directed sparse Hopfield networks.

Many real-world networks are directed, sparse, and hierarchical, with a mixture of feedforward and feedback connections with respect to the hierarchy. Moreover, a small number of master nodes are often able to drive the whole system. We study the dynamics of pattern presentation and recovery on sparse, directed, Hopfield-like neural networks using trophic analysis to characterize their hierarchical structure. This is a recent method which quantifies the local position of each node in a hierarchy (trophic level) as well as the global directionality of the network (trophic coherence). We show that even in a recurrent network, the state of the system can be controlled by a small subset of neurons which can be identified by their low trophic levels. We also find that performance at the pattern recovery task can be significantly improved by tuning the trophic coherence and other topological properties of the network. This may explain the relatively sparse and coherent structures observed in the animal brain and provide insights for improving the architectures of artificial neural networks. Moreover, we expect that the principles we demonstrate here, through numerical analysis, will be relevant for a broad class of system whose underlying network structure is directed and sparse, such as biological, social, or financial networks.

A robust and interpretable machine learning approach using multimodal biological data to predict future pathological tau accumulation.

The early stages of Alzheimer's disease (AD) involve interactions between multiple pathophysiological processes. Although these processes are well studied, we still lack robust tools to predict individualised trajectories of disease progression. Here, we employ a robust and interpretable machine learning approach to combine multimodal biological data and predict future pathological tau accumulation. In particular, we use machine learning to quantify interactions between key pathological markers (ß-amyloid, medial temporal lobe  atrophy, tau and APOE 4) at mildly impaired and asymptomatic stages of AD. Using baseline non-tau markers we derive a prognostic index that: (a) stratifies patients based on future pathological tau accumulation, (b) predicts individualised regional future rate of tau accumulation, and (c) translates predictions from deep phenotyping patient cohorts to cognitively normal individuals. Our results propose a robust approach for fine scale stratification and prognostication with translation impact for clinical trial design targeting the earliest stages of AD.

Establishing a quantitative fluorescence assay for the rapid detection of kynurenine in urine.

The kynurenine metabolite is associated with many diseases and disorders, ranging from diabetes and sepsis to more recently COVID-19. Here we report a fluorescence-based assay for the detection of kynurenine in urine using a specific chemosensor, 3-formyl-4-(ethylthio)-7-(diethylamino)-coumarin. The assay produces a linear response at clinically relevant ranges (1-20 µM), with a limit of detection of 0.7 µM. The average standard addition recoveries of kynurenine in synthetic urine samples are near to 100%, and the relative standard deviation values are less than 8%. The established fluorescence assay for quantitative analysis of kynurenine in urine is facile, sensitive and accurate and holds great potential for low-cost and high-throughput analysis of kynurenine in clinical laboratory settings.

Manifold Alignment Aware Ants: A Markovian Process for Manifold Extraction.

The presence of manifolds is a common assumption in many applications, including astronomy and computer vision. For instance, in astronomy, low-dimensional stellar structures, such as streams, shells, and globular clusters, can be found in the neighborhood of big galaxies such as the Milky Way. Since these structures are often buried in very large data sets, an algorithm, which can not only recover the manifold but also remove the background noise (or outliers), is highly desirable. While other works try to recover manifolds either by pushing all points toward manifolds or by downsampling from dense regions, aiming to solve one of the problems, they generally fail to suppress the noise on manifolds and remove background noise simultaneously. Inspired by the collective behavior of biological ants in food-seeking process, we propose a new algorithm that employs several random walkers equipped with a local alignment measure to detect and denoise manifolds. During the walking process, the agents release pheromone on data points, which reinforces future movements. Over time the pheromone concentrates on the manifolds, while it fades in the background noise due to an evaporation procedure. We use the Markov chain (MC) framework to provide a theoretical analysis of the convergence of the algorithm and its performance. Moreover, an empirical analysis, based on synthetic and real-world data sets, is provided to demonstrate its applicability in different areas, such as improving the performance of t-distributed stochastic neighbor embedding (t-SNE) and spectral clustering using the underlying MC formulas, recovering astronomical low-dimensional structures, and improving the performance of the fast Parzen window density estimator.

Input-to-State Representation in Linear Reservoirs Dynamics.

Reservoir computing is a popular approach to design recurrent neural networks, due to its training simplicity and approximation performance. The recurrent part of these networks is not trained (e.g., via gradient descent), making them appealing for analytical studies by a large community of researchers with backgrounds spanning from dynamical systems to neuroscience. However, even in the simple linear case, the working principle of these networks is not fully understood and their design is usually driven by heuristics. A novel analysis of the dynamics of such networks is proposed, which allows the investigator to express the state evolution using the controllability matrix. Such a matrix encodes salient characteristics of the network dynamics; in particular, its rank represents an input-independent measure of the memory capacity of the network. Using the proposed approach, it is possible to compare different reservoir architectures and explain why a cyclic topology achieves favorable results as verified by practitioners.

Probabilistic learning vector quantization on manifold of symmetric positive definite matrices.

In this paper, we develop a new classification method for manifold-valued data in the framework of probabilistic learning vector quantization. In many classification scenarios, the data can be naturally represented by symmetric positive definite matrices, which are inherently points that live on a curved Riemannian manifold. Due to the non-Euclidean geometry of Riemannian manifolds, traditional Euclidean machine learning algorithms yield poor results on such data. In this paper, we generalize the probabilistic learning vector quantization algorithm for data points living on the manifold of symmetric positive definite matrices equipped with Riemannian natural metric (affine-invariant metric). By exploiting the induced Riemannian distance, we derive the probabilistic learning Riemannian space quantization algorithm, obtaining the learning rule through Riemannian gradient descent. Empirical investigations on synthetic data, image data , and motor imagery electroencephalogram (EEG) data demonstrate the superior performance of the proposed method.

Generalized Learning Riemannian Space Quantization: A Case Study on Riemannian Manifold of SPD Matrices.

Learning vector quantization (LVQ) is a simple and efficient classification method, enjoying great popularity. However, in many classification scenarios, such as electroencephalogram (EEG) classification, the input features are represented by symmetric positive-definite (SPD) matrices that live in a curved manifold rather than vectors that live in the flat Euclidean space. In this article, we propose a new classification method for data points that live in the curved Riemannian manifolds in the framework of LVQ. The proposed method alters generalized LVQ (GLVQ) with the Euclidean distance to the one operating under the appropriate Riemannian metric. We instantiate the proposed method for the Riemannian manifold of SPD matrices equipped with the Riemannian natural metric. Empirical investigations on synthetic data and real-world motor imagery EEG data demonstrate that the performance of the proposed generalized learning Riemannian space quantization can significantly outperform the Euclidean GLVQ, generalized relevance LVQ (GRLVQ), and generalized matrix LVQ (GMLVQ). The proposed method also shows competitive performance to the state-of-the-art methods on the EEG classification of motor imagery tasks.

Positive and negative parenting in conduct disorder with high versus low levels of callous-unemotional traits.

Less is known about the relationship between conduct disorder (CD), callous-unemotional (CU) traits, and positive and negative parenting in youth compared to early childhood. We combined traditional univariate analyses with a novel machine learning classifier (Angle-based Generalized Matrix Learning Vector Quantization) to classify youth (N = 756; 9-18 years) into typically developing (TD) or CD groups with or without elevated CU traits (CD/HCU, CD/LCU, respectively) using youth- and parent-reports of parenting behavior. At the group level, both CD/HCU and CD/LCU were associated with high negative and low positive parenting relative to TD. However, only positive parenting differed between the CD/HCU and CD/LCU groups. In classification analyses, performance was best when distinguishing CD/HCU from TD groups and poorest when distinguishing CD/HCU from CD/LCU groups. Positive and negative parenting were both relevant when distinguishing CD/HCU from TD, negative parenting was most relevant when distinguishing between CD/LCU and TD, and positive parenting was most relevant when distinguishing CD/HCU from CD/LCU groups. These findings suggest that while positive parenting distinguishes between CD/HCU and CD/LCU, negative parenting is associated with both CD subtypes. These results highlight the importance of considering multiple parenting behaviors in CD with varying levels of CU traits in late childhood/adolescence.

Data extraction for epidemiological research (DExtER): a novel tool for automated clinical epidemiology studies.

The use of primary care electronic health records for research is abundant. The benefits gained from utilising such records lies in their size, longitudinal data collection and data quality. However, the use of such data to undertake high quality epidemiological studies, can lead to significant challenges particularly in dealing with misclassification, variation in coding and the significant effort required to pre-process the data in a meaningful format for statistical analysis. In this paper, we describe a methodology to aid with the extraction and processing of such databases, delivered by a novel software programme; the "Data extraction for epidemiological research" (DExtER). The basis of DExtER relies on principles of extract, transform and load processes. The tool initially provides the ability for the healthcare dataset to be extracted, then transformed in a format whereby data is normalised, converted and reformatted. DExtER has a user interface designed to obtain data extracts specific to each research question and observational study design. There are facilities to input the requirements for; eligible study period, definition of exposed and unexposed groups, outcome measures and important baseline covariates. To date the tool has been utilised and validated in a multitude of settings. There have been over 35 peer-reviewed publications using the tool, and DExtER has been implemented as a validated public health surveillance tool for obtaining accurate statistics on epidemiology of key morbidities. Future direction of this work will be the application of the framework to linked as well as international datasets and the development of standardised methods for conducting electronic pre-processing and extraction from datasets for research purposes.

A predictable conserved DNA base composition signature defines human core DNA replication origins.

DNA replication initiates from multiple genomic locations called replication origins. In metazoa, DNA sequence elements involved in origin specification remain elusive. Here, we examine pluripotent, primary, differentiating, and immortalized human cells, and demonstrate that a class of origins, termed core origins, is shared by different cell types and host ~80% of all DNA replication initiation events in any cell population. We detect a shared G-rich DNA sequence signature that coincides with most core origins in both human and mouse genomes. Transcription and G-rich elements can independently associate with replication origin activity. Computational algorithms show that core origins can be predicted, based solely on DNA sequence patterns but not on consensus motifs. Our results demonstrate that, despite an attributed stochasticity, core origins are chosen from a limited pool of genomic regions. Immortalization through oncogenic gene expression, but not normal cellular differentiation, results in increased stochastic firing from heterochromatin and decreased origin density at TAD borders.