Search

Disruptions of axonal connectivity are thought to be a core pathophysiological feature of psychotic illness, but whether they are present early in the illness, prior to antipsychotic exposure, and whether they can predict clinical outcome remain unknown. We acquired diffusion-weighted magnetic resonance images to map structural connectivity between each pair of 319 parcellated brain regions in 61 antipsychotic-naïve individuals with first-episode psychosis (15–25 years, 46% female) and a demographically matched sample of 27 control participants. Clinical follow-up data were also acquired in patients 3 and 12 months after the scan. We used connectome-wide analyses to map disruptions of inter-regional pairwise connectivity and connectome-based predictive modeling to predict longitudinal change in symptoms and functioning. Individuals with first-episode psychosis showed disrupted connectivity in a brainwide network linking all brain regions compared with controls (familywise error–corrected p =.03). Baseline structural connectivity significantly predicted change in functioning over 12 months (r = 0.44, familywise error–corrected p =.041), such that lower connectivity within fronto-striato-thalamic systems predicted worse functional outcomes. Brainwide reductions of structural connectivity exist during the early stages of psychotic illness and cannot be attributed to antipsychotic medication. Moreover, baseline measures of structural connectivity can predict change in patient functional outcomes up to 1 year after engagement with treatment services. [ABSTRACT FROM AUTHOR]

Multimodal data analysis has attracted ever-increasing attention in computational biology and bioinformatics community recently. However, existing multimodal learning approaches need all data modalities available at both training and prediction stages, thus they cannot be applied to many real-world biomedical applications, which often have a missing modality problem as the collection of all modalities is prohibitively costly. Meanwhile, two diagnosis-related pieces of information are of main interest during the examination of a subject regarding a chronic disease (with longitudinal progression): their current status (diagnosis) and how it will change before next visit (longitudinal outcome). Correct responses to these queries can identify susceptible individuals and provide the means of early interventions for them. In this article, we develop a novel adversarial mutual learning framework for longitudinal disease progression prediction, allowing us to leverage multiple data modalities available for training to train a performant model that uses a single modality for prediction. Specifically, in our framework, a single-modal model (which utilizes the main modality) learns from a pretrained multimodal model (which accepts both main and auxiliary modalities as input) in a mutual learning manner to (1) infer outcome-related representations of the auxiliary modalities based on its own representations for the main modality during adversarial training and (2) successfully combine them to predict the longitudinal outcome. We apply our method to analyze the retinal imaging genetics for the early diagnosis of age-related macular degeneration (AMD) disease, that is, simultaneous assessment of the severity of AMD at the time of the current visit and the prognosis of the condition at the subsequent visit. Our experiments using the Age-Related Eye Disease Study dataset show that our method is more effective than baselines at classifying patients' current and forecasting their future AMD severity. [ABSTRACT FROM AUTHOR]

The long-term negative consequences of learning difficulties have been acknowledged. Nonetheless, research is still scarce regarding the prediction of adolescent difficulties in reading and arithmetic skills. The present study examines at which age phase and with what kind of constellation of parent- and child-related factors can adolescent difficulties in arithmetic and/or reading fluency be successfully predicted. A sample of Finnish children (N = 941) was followed from the onset of kindergarten (at age 6) through adolescence (ages 13–16). Children’s cognitive skills were assessed in kindergarten, and arithmetic and reading fluency were examined in Grades 2, 4, 6, 7, and 9. Parents’ self-report data were collected on their own learning difficulties and educational level. Scoring below the 16th percentile in both Grades 7 and 9 was set as the criterion for dysfluency either in reading (N = 87, 9.2%) or arithmetic (N = 84, 8.9%). Adolescent dysfluency in both domains was moderately predicted by parental measures and kindergarten cognitive skills. Although adding school-age fluency measures clearly increased both the predictability and specificity of models up to Grade 4 for both skills, knowledge of letters’ names, counting, and visuospatial skills remained unique predictors of dysfluency in adolescence. [ABSTRACT FROM AUTHOR]

This study investigated endogenous and exogenous predictors of early literacy in Turkish-speaking children. Whether children's language and working memory performances (as the endogenous factors) and home literacy environment (as the exogenous factor) in the beginning of kindergarten predict the children's current and year-end early literacy skills (phonological awareness, letter knowledge, receptive and expressive vocabulary) was examined. The participants consisted of 441 kindergarten children. Results showed that language development, working memory, and home reading environment predicted children's both current and year-end phonological awareness. Language and home writing activities were significant predictors of the year-end letter knowledge. Working memory was a significant predictor for both the current and year-end letter knowledge. Language, working memory and home reading environment significantly predicted the acquisition of receptive and expressive vocabulary. In conclusion, results suggest that each of the early literacy skills is related to both the developmental characteristics of children and their home literacy environment. [ABSTRACT FROM AUTHOR]

Purpose: Radiotherapy presents unique challenges and clinical requirements for longitudinal tumor and organ‐at‐risk (OAR) prediction during treatment. The challenges include tumor inflammation/edema and radiation‐induced changes in organ geometry, whereas the clinical requirements demand flexibility in input/output sequence timepoints to update the predictions on rolling basis and the grounding of all predictions in relationship to the pre‐treatment imaging information for response and toxicity assessment in adaptive radiotherapy. Methods: To deal with the aforementioned challenges and to comply with the clinical requirements, we present a novel 3D sequence‐to‐sequence model based on Convolution Long Short‐Term Memory (ConvLSTM) that makes use of series of deformation vector fields (DVFs) between individual timepoints and reference pre‐treatment/planning CTs to predict future anatomical deformations and changes in gross tumor volume as well as critical OARs. High‐quality DVF training data are created by employing hyper‐parameter optimization on the subset of the training data with DICE coefficient and mutual information metric. We validated our model on two radiotherapy datasets: a publicly available head‐and‐neck dataset (28 patients with manually contoured pre‐, mid‐, and post‐treatment CTs), and an internal non‐small cell lung cancer dataset (63 patients with manually contoured planning CT and 6 weekly CBCTs). Results: The use of DVF representation and skip connections overcomes the blurring issue of ConvLSTM prediction with the traditional image representation. The mean and standard deviation of DICE for predictions of lung GTV at weeks 4, 5, and 6 were 0.83 ± 0.09, 0.82 ± 0.08, and 0.81 ± 0.10, respectively, and for post‐treatment ipsilateral and contralateral parotids, were 0.81 ± 0.06 and 0.85 ± 0.02. Conclusion: We presented a novel DVF‐based Seq2Seq model for medical images, leveraging the complete 3D imaging information of a relatively large longitudinal clinical dataset, to carry out longitudinal GTV/OAR predictions for anatomical changes in HN and lung radiotherapy patients, which has potential to improve RT outcomes. [ABSTRACT FROM AUTHOR]

This 17-year prospective study applied a social-developmental lens to the challenge of distinguishing predictors of adolescent-era substance use from predictors of longer term adult substance use problems. A diverse community sample of 168 individuals was repeatedly assessed from age 13 to age 30 using test, self-, parent-, and peer-report methods. As hypothesized, substance use within adolescence was linked to a range of likely transient social and developmental factors that are particularly salient during the adolescent era, including popularity with peers, peer substance use, parent–adolescent conflict, and broader patterns of deviant behavior. Substance abuse problems at ages 27–30 were best predicted, even after accounting for levels of substance use in adolescence, by adolescent-era markers of underlying deficits, including lack of social skills and poor self-concept. The factors that best predicted levels of adolescent-era substance use were not generally predictive of adult substance abuse problems in multivariate models (either with or without accounting for baseline levels of use). Results are interpreted as suggesting that recognizing the developmental nature of adolescent-era substance use may be crucial to distinguishing factors that predict socially driven and/or relatively transient use during adolescence from factors that predict long-term problems with substance abuse that extend well into adulthood. [ABSTRACT FROM AUTHOR]

• Developed an accurate prediction system for the progression of mild cognitive impairment to AD. • Used longitudinal structural MRI data without the segmentation of regions of interest. • Applied dictionary learning to exploit the nuances between mild cognitive impairment patients. Efficient prediction of the progression of mild cognitive impairment (MCI) to Alzheimer's disease (AD) is important for the early intervention and management of AD. The aim of our study was to develop a longitudinal structural magnetic resonance imaging-based prediction system for MCI progression. A total of 164 MCI patients with longitudinal data were collected from the Alzheimer's Disease Neuroimaging Initiative (ADNI). After preprocessing, a discriminative dictionary learning framework was applied to differentiate MCI patches, avoiding the segmentation of regions of interest. Then, the proportion of patches classified as more severe atrophy patches in a patient was calculated as his or her feature to be input into a simple support vector machine. Finally, a new subject was predicted with fourfold cross-validation (CV), and the area under the receiver operating characteristic curve (AUC) was determined. The average accuracy and AUC values after fourfold CV were 0.973 and 0.984, respectively. The effects of the data from one or two time points were also investigated. The proposed prediction system achieves desirable and reliable performance in predicting progression for MCI patients. Additionally, the prediction of MCI progression with longitudinal data was more effective and accurate. The developed scheme is expected to advance the clinical research and treatment of MCI patients. [ABSTRACT FROM AUTHOR]

Missing data is a common problem in longitudinal studies due to subject dropouts and failed scans. We present a graph-based convolutional neural network to predict missing diffusion MRI data. In particular, we consider the relationships between sampling points in the spatial domain and the diffusion wave-vector domain to construct a graph. We then use a graph convolutional network to learn the non-linear mapping from available data to missing data. Our method harnesses a multi-scale residual architecture with adversarial learning for prediction with greater accuracy and perceptual quality. Experimental results show that our method is accurate and robust in the longitudinal prediction of infant brain diffusion MRI data. [ABSTRACT FROM AUTHOR]

Attention deficit hyperactivity disorder (ADHD) is emblematic of unresolved heterogeneity in psychiatric disorders-the variation in biological, clinical, and psychological correlates that impedes progress on etiology. One approach to this problem is to characterize subgroups using measures rooted in biological or psychological theory, consistent with the National Institute of Mental Health's research domain criteria initiative. Within ADHD, a promising application involves using emotion trait profiles that can address the role of irritability as a complicating feature for ADHD. Here, a new sample of 186 children with ADHD was evaluated using community detection analysis to determine if meaningful subprofiles existed and if they replicated those previously identified. The new sample and a prior sample were pooled for evaluation of (a) method dependence, (b) longitudinal assessment of the stability of classifications, and (c) clinical prediction 2 years later. Three temperament profiles were confirmed within the ADHD group: one with normative emotional functioning ("mild"), one with high surgency ("surgent"), and one with high negative affect ("irritable"). Profiles were similar across statistical clustering approaches. The irritable group had the highest external validity: It was moderately stable over time and it enhanced prospective prediction of clinical outcomes beyond standard baseline indicators. The irritable group was not reducible to ADHD + oppositional defiant disorder, ADHD + disruptive mood dysregulation disorder, or other patterns of comorbidity. Among the negative affect domains studied, trait proneness to anger uniquely contributed to clinical prediction. Results extend our understanding of chronic irritability in psychiatric disorders and provide prospects for a fresh approach to assessing ADHD heterogeneity focused on the distinction between ADHD with and without anger/irritability. (PsycINFO Database Record (c) 2019 APA, all rights reserved). [ABSTRACT FROM AUTHOR]

Cross-sectional studies have established that performance on paired-associate learning (PAL) tasks is associated with reading performance. There are good reasons to expect such a relationship because learning to read involves learning the sounds of individual letters and possibly also sounds of strings of letters (e.g., spelling patterns). However, results from longitudinal studies have been mixed. A closer look at these studies suggests that PAL may be related to development of accuracy rather than speed in reading. This suggestion was investigated directly in the current longitudinal study. The study followed 137 students from Grade 0 (kindergarten) to Grade 5. In Grade 0, they completed measures of PAL, letter knowledge, phoneme awareness, and rapid automatized naming (RAN). In Grades 1 and 5, decoding accuracy was measured with the addition of decoding speed in Grade 5. PAL in Grade 0 was found to be a unique predictor of decoding accuracy in Grades 1 and 5 after controlling for Grade 0 letter knowledge, phoneme awareness, and RAN. PAL in Grade 0 even contributed to Grade 5 decoding accuracy after also controlling for Grade 1 decoding. Zero-order correlations between PAL and Grade 5 decoding speed were nonsignificant and close to zero. The results indicate that PAL measures a trait that may influence reading development over a substantial amount of time. Possible roles of PAL in decoding development over time are discussed, for example, how verbal learning may be a core component in the acquisition of associations between letter patterns (spelling patterns) and their pronunciation. [ABSTRACT FROM AUTHOR]

• We propose a trustworthy adversarial learning metamorphosis framework that accounts for both the appearance and structural changes in infant brain MRI. • A spatial-frequency transfer block captures structural changes from multiple frequency bands with wavelet decomposition. • A quality-guided learning strategy is incorporated to improve predictions in challenging regions. • A multi-scale hybrid loss function is employed to improve prediction of textural details and anatomical boundaries. Missing scans are inevitable in longitudinal studies due to either subject dropouts or failed scans. In this paper, we propose a deep learning framework to predict missing scans from acquired scans, catering to longitudinal infant studies. Prediction of infant brain MRI is challenging owing to the rapid contrast and structural changes particularly during the first year of life. We introduce a trustworthy metamorphic generative adversarial network (MGAN) for translating infant brain MRI from one time point to another. MGAN has three key features: (i) Image translation leveraging spatial and frequency information for detail-preserving mapping; (ii) Quality-guided learning strategy that focuses attention on challenging regions. (iii) Multi-scale hybrid loss function that improves translation of image contents. Experimental results indicate that MGAN outperforms existing GANs by accurately predicting both tissue contrasts and anatomical details. [ABSTRACT FROM AUTHOR]

As connected autonomous vehicles (CAVs) enter the fleet, there will be a long period when these vehicles will have to interact with human drivers. One of the challenges for CAVs is that human drivers do not communicate their decisions well. Fortunately, the kinematic behavior of a human-driven vehicle may be a good predictor of driver intent within a short time frame. We analyzed the kinematic time series data (e.g., speed) for a set of drivers making left turns at intersections to predict whether the driver would stop before executing the turn. We used principal components analysis (PCA) to generate independent dimensions that explain the variation in vehicle speed before a turn. These dimensions remained relatively consistent throughout the maneuver, allowing us to compute independent scores on these dimensions for different time windows throughout the approach to the intersection. We then linked these PCA scores to whether a driver would stop before executing a left turn using the random intercept Bayesian additive regression trees. Five more road and observable vehicle characteristics were included to enhance prediction. Our model achieved an area under the receiver operating characteristic curve (AUC) of 0.84 at 94 m away from the center of an intersection and steadily increased to 0.90 by 46 m away from the center of an intersection. [ABSTRACT FROM AUTHOR]

Objective: The hostile-world scenario (HWS) denotes a personal belief system regarding threats to one’s physical and mental integrity. We examined whether the HWS predicted health among older adults. Method: The Israeli branch of the Survey of Health, Ageing and Retirement in Europe (SHARE-Israel) provided data on 1,286 participants, aged 50+, interviewed in two waves 4 years apart. A special measure assembled items pertinent to the HWS throughout the SHARE survey. Nine outcomes indicated physical health (e.g., activities of daily living, medical conditions) and mental health (e.g., depressive symptoms, satisfaction with life). Results: The HWS at Wave 1 predicted all physical and mental outcomes at Wave 2, except cognitive functioning, beyond effects of sociodemographics and the respective outcome’s baseline at Wave 1. This predictive effect was stronger among older participants. Discussion: The results support the conception of the HWS as a psychological monitor that senses approaching functional declines in later life. [ABSTRACT FROM AUTHOR]

In this longitudinal study, we examined teachers' self-efficacy as a long-term predictor of their mastery goal orientation and three dimensions of instructional quality: supportive classroom climate, effective classroom management, and cognitive activation. Mastery goal orientation was also analyzed as a predictor of instructional quality. Teachers' optimism, engagement, and strain were assessed to gain information about the construct validity of the scales on self-efficacy and mastery goal orientation. We analyzed the self-report data of 203 German in-service teachers who participated in all of three time points of assessment (the years 2001, 2008, and 2011). Confirmatory factor analyses supported the assumed three-dimensionality of instructional quality. Teacher self-efficacy was found to be relatively stable and to be a long-term predictor of instructional quality as indicated by the results of latent variable modeling. Moreover, instructional quality is predicted by mastery goal orientation, which in turn is regressed on self-efficacy. As supported also by bias-corrected bootstrapping, mastery goal orientation partially mediated the relationship between classroom climate and self-efficacy. Results and an outlook for future research are discussed. [ABSTRACT FROM AUTHOR]

The cross-sectional and longitudinal influences of positive youth development and life satisfaction on adolescent problem behaviour were examined in this study. The respondents were 4,523 Chinese students recruited from 43 secondary schools in Hong Kong, and the study followed them from Grade 7, Grade 9 to Grade 11. Repeated measurements employing validated self-reported assessment tools were used. Analyses using structural equation modelling showed that positive youth development directly influenced life satisfaction and reduced problem behaviour at Grades 7 and 9. However, the direct effect of positive youth development on problem behaviour was mediated by life satisfaction at Grade 11. The mediating effect of life satisfaction on the enhancement of future positive youth development was also discovered. These pioneer longitudinal findings contribute to the existing literature in delineating the causal and mediating roles of positive youth development and life satisfaction in adolescent problem behaviour, particularly in the Chinese context. [ABSTRACT FROM AUTHOR]

Longitudinal predictors of dietary behavior change are important and in need of study. This secondary data analysis combined primary data across three randomized trials to examine transtheoretical model (TTM) and specific dietary predictors of successful dietary change at 12 and 24 months separately in treatment and control groups (N = 4178). The treatment group received three TTM-tailored print interventions over 12 months between 1995 and 2000. Chi-square and MANOVA analyses were used to examine baseline predictors of dietary outcome at 12 and 24 months. Last, a multivariable logistic regression was conducted with all baseline variables included. Across all analyses in both treatment and control groups, the most robust predictors of successful change were for TTM-tailored treatment group, preparation stage of change, and increased use of dietary behavior variables such as moderating fat intake, substitution of lower fat foods, and increasing intake of healthful foods. These results provide strong evidence for treatment, stage and behavioral dietary severity effects predicting dietary behavior change over time, and for targeting these variables with the strongest relationships to outcome in interventions, such as TTM-tailored dietary interventions. [ABSTRACT FROM AUTHOR]

Abstract: Longitudinal predictors of dietary behavior change are important and in need of study. This secondary data analysis combined primary data across three randomized trials to examine transtheoretical model (TTM) and specific dietary predictors of successful dietary change at 12 and 24months separately in treatment and control groups (N=4178). The treatment group received three TTM-tailored print interventions over 12months between 1995 and 2000. Chi-square and MANOVA analyses were used to examine baseline predictors of dietary outcome at 12 and 24months. Last, a multivariable logistic regression was conducted with all baseline variables included. Across all analyses in both treatment and control groups, the most robust predictors of successful change were for TTM-tailored treatment group, preparation stage of change, and increased use of dietary behavior variables such as moderating fat intake, substitution of lower fat foods, and increasing intake of healthful foods. These results provide strong evidence for treatment, stage and behavioral dietary severity effects predicting dietary behavior change over time, and for targeting these variables with the strongest relationships to outcome in interventions, such as TTM-tailored dietary interventions. [Copyright &y& Elsevier]

Background/Aims: To investigate the longitudinal stability and progression of different subtypes of mild cognitive impairment (MCI) in older adults. Methods: We classified 217 individuals with no cognitive impairment (NCI), amnestic MCI (aMCI) based on a single test (aMCI-1) or multiple tests (aMCI-2+), nonamnestic MCI (naMCI) based on a single test (naMCI-1) or multiple tests (naMCI-2+), or amnestic + nonamnestic MCI (a+naMCI), using their baseline neuropsychological test scores, and performed annual follow-up evaluations for up to 3 years. Results: None of the subjects with aMCI-2+ reverted to normal during follow-up, with 50% of these subjects remaining stable and 50% worsening over time. Similarly, less than 20% of subjects with aMCI-2+ and a+naMCI reverted to NCI during the follow-up period, whereas 50% of aMCI-1 and 37% with naMCI-1 reverted to NCI during this same period. Conclusion: Reversion to NCI occurs much more frequently when the diagnosis of MCI is based on the results of a single neuropsychological test than when it is based on the results of more memory tests. In epidemiological studies and clinical trials the diagnosis of MCI will likely be more stable if impairment on more than one test is required for amnestic and/or nonamnestic domains. Copyright © 2009 S. Karger AG, Basel [ABSTRACT FROM AUTHOR]

Development of the psychosomatic problems picky eating and headache and stomachache complaints in middle childhood was investigated from an emotion regulation perspective. The role of negative emotionality and family emotion regulatory factors (attachment to mother and parental perceived control) was studied. The sample (N=87) was a predominantly middleclass, community sample. The study was longitudinal, based on data from several data collections between child age 11 months and 9 years. The results showed that headache and stomachache complaints were mainly predicted by early negative emotionality, and picky eating by the family factors. More negative emotionality, insecure attachment and less perceived control were related to more psychosomatic problems in linear and interaction models. The findings were interpreted as showing that by considering emotion regulation, a fruitful perspective for understanding the development of psychosomatic problems could be elaborated. Copyright © 2004 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]

Can vision tests predict subsequent loss of acuity? The association between performance on several low contrast spatial vision measures, glare recovery, color discrimination, flicker sensitivity, stereopsis and ocular disease status at baseline and acuity loss 4.4 years later was examined in a large aged random sample with good initial acuity. In univariate analyses, several vision measures, retinal disease status and age were each significant predictors of subsequent acuity loss. In a multiple regression analysis, only low contrast spatial vision was a significant predictor, but the other vision measures, retinal disease status and age were not. For each doubling of low contrast spatial vision threshold at baseline, individuals were more than two times as likely to suffer subsequent significant visual acuity loss. Tests of low contrast spatial vision are strong predictors of significant subsequent visual acuity loss. These findings have implications for clinical trials, clinical management, and acceptance of these measures into clinical practice. [ABSTRACT FROM AUTHOR]

The 4-year stability in problem behaviors assessed with the Child Behavior Checklist (CBCL), and their predictive validity with regard to poor outcome variables was studied in 1052 4-12-year old children from a general population sample. Irrespective of sex and age, the scores on [he CBCL of some 44% of the children remained above the 90th percentile over the 4-year time interval. Externalizing problem behavior tended to show somewhat greater stability than internalizing problem behavior. High initial levels of both internalizing and externalizing problems, and persistence of problems were predictive of referral to mental health services. Referral to special education was higher for younger boys with high initial attention problems and school problems. Police contacts occurred more often in older boys with high initial aggression scores. [ABSTRACT FROM AUTHOR]