A distributed multitask multimodal approach for the prediction of Alzheimer's disease in a longitudinal study.

Predicting the progression of Alzheimer's Disease (AD) has been held back for decades due to the lack of sufficient longitudinal data required for the development of novel machine learning algorithms. This study proposes a novel machine learning algorithm for predicting the progression of Alzheimer's disease using a distributed multimodal, multitask learning method. More specifically, each individual task is defined as a regression model, which predicts cognitive scores at a single time point. Since the prediction tasks for multiple intervals are related to each other in chronological order, multitask regression models have been developed to track the relationship between subsequent tasks. Furthermore, since subjects have various combinations of recording modalities together with other genetic, neuropsychological and demographic risk factors, special attention is given to the fact that each modality may experience a specific sparsity pattern. The model is hence generalized by exploiting multiple individual multitask regression coefficient matrices for each modality. The outcome for each independent modality-specific learner is then integrated with complementary information, known as risk factor parameters, revealing the most prevalent trends of the multimodal data. This new feature space is then used as input to the gradient boosting kernel in search for a more accurate prediction. This proposed model not only captures the complex relationships between the different feature representations, but it also ignores any unrelated information which might skew the regression coefficients. Comparative assessments are made between the performance of the proposed method with several other well-established methods using different multimodal platforms. The results indicate that by capturing the interrelatedness between the different modalities and extracting only relevant information in the data, even in an incomplete longitudinal dataset, will yield minimized prediction errors.

Scenic routing navigation using property valuation.

Extensive prior work has provided methods for the optimization of routing based on weights assigned to travel duration, and/or travel cost, and/or the distance traveled. Routing can be in various modalities, such as by car, on foot, by bicycle, via public transit, or by boat. A typical method of routing involves building a graph comprised of street segments, assigning a normalized weighted value to each segment, and then applying the weighted-shorted path algorithm to the graph in order to find the best route. Some users desire that the routing suggestion include consideration pertaining to the scenic-architectural quality of the path. For example, a user may seek a leisure walk via what they might deem as visually attractive architecture. Here, we are proposing a method to quantify such user preferences and scenic quality and to augment the standard routing methods by giving weight to the scenic quality. That is, instead of suggesting merely the time and cost-optimal route, we will find the best route that is tailored towards the user's scenic quality preferences as an additional criterion to the time and cost. The proposed method uniquely weighs the scenic interest or residential street segments based on the property valuation data.

A Tensorized Multitask Deep Learning Network for Progression Prediction of Alzheimer's Disease.

With the advances in machine learning for the diagnosis of Alzheimer's disease (AD), most studies have focused on either identifying the subject's status through classification algorithms or on predicting their cognitive scores through regression methods, neglecting the potential association between these two tasks. Motivated by the need to enhance the prospects for early diagnosis along with the ability to predict future disease states, this study proposes a deep neural network based on modality fusion, kernelization, and tensorization that perform multiclass classification and longitudinal regression simultaneously within a unified multitask framework. This relationship between multiclass classification and longitudinal regression is found to boost the efficacy of the final model in dealing with both tasks. Different multimodality scenarios are investigated, and complementary aspects of the multimodal features are exploited to simultaneously delineate the subject's label and predict related cognitive scores at future timepoints using baseline data. The main intent in this multitask framework is to consolidate the highest accuracy possible in terms of precision, sensitivity, F1 score, and area under the curve (AUC) in the multiclass classification task while maintaining the highest similarity in the MMSE score as measured through the correlation coefficient and the RMSE for all time points under the prediction task, with both tasks, run simultaneously under the same set of hyperparameters. The overall accuracy for multiclass classification of the proposed KTMnet method is 66.85 ± 3.77. The prediction results show an average RMSE of 2.32 ± 0.52 and a correlation of 0.71 ± 5.98 for predicting MMSE throughout the time points. These results are compared to state-of-the-art techniques reported in the literature. A discovery from the multitasking of this consolidated machine learning framework is that a set of hyperparameters that optimize the prediction results may not necessarily be the same as those that would optimize the multiclass classification. In other words, there is a breakpoint beyond which enhancing further the results of one process could lead to the downgrading in accuracy for the other.

Sub-patterns of language network reorganization in pediatric localization related epilepsy: a multisite study.

To study the neural networks reorganization in pediatric epilepsy, a consortium of imaging centers was established to collect functional imaging data. Common paradigms and similar acquisition parameters were used. We studied 122 children (64 control and 58 LRE patients) across five sites using EPI BOLD fMRI and an auditory description decision task. After normalization to the MNI atlas, activation maps generated by FSL were separated into three sub-groups using a distance method in the principal component analysis (PCA)-based decisional space. Three activation patterns were identified: (1) the typical distributed network expected for task in left inferior frontal gyrus (Broca's) and along left superior temporal gyrus (Wernicke's) (60 controls, 35 patients); (2) a variant left dominant pattern with greater activation in IFG, mesial left frontal lobe, and right cerebellum (three controls, 15 patients); and (3) activation in the right counterparts of the first pattern in Broca's area (one control, eight patients). Patients were over represented in Groups 2 and 3 (P < 0.0004). There were no scanner (P = 0.4) or site effects (P = 0.6). Our data-driven method for fMRI activation pattern separation is independent of a priori notions and bias inherent in region of interest and visual analyses. In addition to the anticipated atypical right dominant activation pattern, a sub-pattern was identified that involved intensity and extent differences of activation within the distributed left hemisphere language processing network. These findings suggest a different, perhaps less efficient, cognitive strategy for LRE group to perform the task.

Dynamics and Distant Effects of Frontal/Temporal Epileptogenic Focus Using Functional Connectivity Maps.

OBJECTIVE: Connectivity patterns of interictal epileptiform discharges are all subtle indicators of where the three-dimensional (3D) source of a seizure could be located. These specific patterns are explored in the recorded electroencephalogram (EEG) signals of 20 individuals diagnosed with focal epilepsy to assess how their functional brain maps could be affected by the 3D onset of a seizure. METHODS: Functional connectivity maps, estimated by phase synchrony among EEG electrodes, were obtained by applying a data-driven recurrence-based method. This is augmented through a novel approach for selecting optimal parameters that produce connectivity matrices that are deemed significant for assessing epileptiform activity in context to the 3D source localization of seizure onset. These functional connectivity matrices were evaluated in different brain areas to gauge the regional effects of the 3D epileptic source. RESULTS: Empirical evaluations indicate high synchronization in the temporal and frontal areas of the effected epileptic hemisphere, whereas strong links connect the irritated area to frontal and temporal lobes of the opposite hemisphere. CONCLUSION: Epileptic activity originating in the temporal or frontal areas is seen to affect these areas in both hemispheres. SIGNIFICANCE: The results obtained express the dynamics of focal epilepsy in context to both the epileptogenic zone and the affected distant areas of the brain.

A Gaussian-based model for early detection of mild cognitive impairment using multimodal neuroimaging.

BACKGROUND: Diagnosis of early mild cognitive impairment (EMCI) as a prodromal stage of Alzheimer's disease (AD) with its delineation from the cognitively normal (CN) group remains a challenging but essential step for the planning of early treatment. Although several studies have focused on the MCI diagnosis, this study introduces the early stage of MCI to assess more thoroughly the earliest signs of disease manifestation and progression. NEW METHOD: We used random forest feature selection model with a Gaussian-based algorithm to perform method evaluation. This integrated method serves to define multivariate normal distributions in order to classify different stages of AD, with the focus placed on detecting EMCI subjects in the most challenging classification of CN vs. EMCI. RESULTS: Using 896 participants classified into the four categories of CN, EMCI, late mild cognitive impairment (LMCI) and AD, the results show that the EMCI group can be delineated from the CN group with a relatively high accuracy of 78.8% and sensitivity of 81.3%. COMPARISON WITH EXISTING METHOD(S): The feature selection model and classifier are compared with some other prominent algorithms. Although higher accuracy has been achieved using the Gaussian process (GP) model (78.8%) over the SVM classifier (75.6%) for CN vs. EMCI classification, with 0.05 being the cutoff for significance, and based on student's t-test, it was determined that the differences for accuracy, sensitivity, specificity between the GP method and support vector machine (SVM) are not statistically significant. CONCLUSION: Addressing the challenging classification of CN vs. EMCI provides useful information to help clinicians and researchers determine essential measures that can help in the early detection of AD.

Gaussian Discriminant Analysis for Optimal Delineation of Mild Cognitive Impairment in Alzheimer's Disease.

Over the past few years, several approaches have been proposed to assist in the early diagnosis of Alzheimer's disease (AD) and its prodromal stage of mild cognitive impairment (MCI). Using multimodal biomarkers for this high-dimensional classification problem, the widely used algorithms include Support Vector Machines (SVM), Sparse Representation-based classification (SRC), Deep Belief Networks (DBN) and Random Forest (RF). These widely used algorithms continue to yield unsatisfactory performance for delineating the MCI participants from the cognitively normal control (CN) group. A novel Gaussian discriminant analysis-based algorithm is thus introduced to achieve a more effective and accurate classification performance than the aforementioned state-of-the-art algorithms. This study makes use of magnetic resonance imaging (MRI) data uniquely as input to two separate high-dimensional decision spaces that reflect the structural measures of the two brain hemispheres. The data used include 190 CN, 305 MCI and 133 AD subjects as part of the AD Big Data DREAM Challenge #1. Using 80% data for a 10-fold cross-validation, the proposed algorithm achieved an average F1 score of 95.89% and an accuracy of 96.54% for discriminating AD from CN; and more importantly, an average F1 score of 92.08% and an accuracy of 90.26% for discriminating MCI from CN. Then, a true test was implemented on the remaining 20% held-out test data. For discriminating MCI from CN, an accuracy of 80.61%, a sensitivity of 81.97% and a specificity of 78.38% were obtained. These results show significant improvement over existing algorithms for discriminating the subtle differences between MCI participants and the CN group.

Using Big Data to Evaluate the Association between Periodontal Disease and Rheumatoid Arthritis.

An association between periodontal disease and rheumatoid arthritis is believed to exist. Most investigations into a possible relationship have been case-control studies with relatively low sample sizes. The advent of very large clinical repositories has created new opportunities for data-driven research. We conducted a retrospective cohort study to measure the association between periodontal disease and rheumatoid arthritis in a population of 25 million patients. We demonstrated that subjects with periodontal disease were roughly 1.4 times more likely to have rheumatoid arthritis. These results compare favorably with those of previous studies on smaller cohorts. Additional work is needed to identify the mechanisms behind this association and to determine if aggressive treatment of periodontal disease can alter the course of rheumatoid arthritis.

Introduction to personalized medicine in diabetes mellitus.

The world is facing an epidemic rise in diabetes mellitus (DM) incidence, which is challenging health funders, health systems, clinicians, and patients to understand and respond to a flood of research and knowledge. Evidence-based guidelines provide uniform management recommendations for "average" patients that rarely take into account individual variation in susceptibility to DM, to its complications, and responses to pharmacological and lifestyle interventions. Personalized medicine combines bioinformatics with genomic, proteomic, metabolomic, pharmacogenomic ("omics") and other new technologies to explore pathophysiology and to characterize more precisely an individual's risk for disease, as well as response to interventions. In this review we will introduce readers to personalized medicine as applied to DM, in particular the use of clinical, genetic, metabolic, and other markers of risk for DM and its chronic microvascular and macrovascular complications, as well as insights into variations in response to and tolerance of commonly used medications, dietary changes, and exercise. These advances in "omic" information and techniques also provide clues to potential pathophysiological mechanisms underlying DM and its complications.

Metabolic profiling in personalized medicine: bridging the gap between knowledge and clinical practice in Type 2 diabetes.

Type 2 diabetes mellitus (DM2) is the most commonly diagnosed metabolic disease and its prevalence is expected to increase. Epidemiological studies clearly show excess mortality associated with DM2, as well as an increased risk of DM2-related complications. Advances in personalized medicine would greatly improve patient care in the field of diabetes and other metabolic diseases. Prediction of the disease in asymptomatic patients as well as its harsh complications in patients already diagnosed is becoming a necessity, with the considerable increase in the cost of the treatment. In the current article, we review the known clinical, molecular metabolic and genetic biomarkers that should be integrated in a future bioinformatic platform to be used at the point-of-care, and discuss the challenges we face in applying this vision of personalized medicine for diabetes into reality.

Measurement of pupil diameter variations as a physiological indicator of the affective state in a computer user.

In this paper we explore the potential of analyzing pupil diameter measurements obtained using a desktop-mounted Eye Gaze Tracking (EGT) instrument to identify changes in the affective state of a computer user. In our experiment we induced intervals of relaxed and stressed affective states by asking the computer user to respond to sequences of congruent and incongruent Stroop word presentations. The recorded pupil diameter values verify our initial expectations by showing an increase in pupil diameter mean value as the subject transitions form a congruent Stroop sequence or segment to an incongruent Stroop segment. This mean pupil diameter value increase was found in all but one of the 96 transitions studied from 32 subjects. The statistical significance of these mean value increases was studied by means of a t-test for the comparison of means. All the pairs of mean pupil diameter values were found to be significantly different at a 0.05 significance level (p <0.05). This seems to indicate that the real-time measurement of pupil diameter of the computer user holds a strong promise to become a non-intrusive way to make the computer aware of changes in the affective status of the user. However, it is clear that several implementation issues still must be resolved before this approach can be practical for use in the context of ordinary human-computer interaction, where, for example, the environmental illumination levels are not controlled, as they were during our experiments.

Heart rate variability biofeedback as a method for assessing baroreflex function: a preliminary study of resonance in the cardiovascular system.

This study describes the use of a biofeedback method for the noninvasive study of baroreflex mechanisms. Five previously untrained healthy male participants learned to control oscillations in heart rate using biofeedback training to modify their heart rate variability at specific frequencies. They were instructed to match computer-generated sinusoidal oscillations with oscillations in heart rate at seven frequencies within the range of 0.01-0.14 Hz. All participants successfully produced high-amplitude target-frequency oscillations in both heart rate and blood pressure. Stable and predictable transfer functions between heart rate and blood pressure were obtained in all participants. The highest oscillation amplitudes were produced in the range of 0.055-0.11 Hz for heart rate and 0.02-0.055 Hz for blood pressure. Transfer functions were calculated among sinusoidal oscillations in the target stimuli, heart rate, blood pressure, and respiration for frequencies at which subjects received training. High and low target-frequency oscillation amplitudes at specific frequencies could be explained by resonance among various oscillatory processes in the cardiovascular system. The exact resonant frequencies differed among individuals. Changes in heart rate oscillations could not be completely explained by changes in breathing. The biofeedback method also allowed us to quantity characteristics of inertia, delay, and speed sensitivity in baroreflex system. We discuss the implications of these findings for using heart rate variability biofeedback as an aid in diagnosing various autonomic and cardiovascular system disorders and as a method for treating these disorders.