skip to main content

Title: EarlyScreen: Multi-scale Instance Fusion for Predicting Neural Activation and Psychopathology in Preschool Children
Emotion dysregulation in early childhood is known to be associated with a higher risk of several psychopathological conditions, such as ADHD and mood and anxiety disorders. In developmental neuroscience research, emotion dysregulation is characterized by low neural activation in the prefrontal cortex during frustration. In this work, we report on an exploratory study with 94 participants aged 3.5 to 5 years, investigating whether behavioral measures automatically extracted from facial videos can predict frustration-related neural activation and differentiate between low- and high-risk individuals. We propose a novel multi-scale instance fusion framework to develop EarlyScreen - a set of classifiers trained on behavioral markers during emotion regulation. Our model successfully predicts activation levels in the prefrontal cortex with an area under the receiver operating characteristic (ROC) curve of 0.85, which is on par with widely-used clinical assessment tools. Further, we classify clinical and non-clinical subjects based on their psychopathological risk with an area under the ROC curve of 0.80. Our model's predictions are consistent with standardized psychometric assessment scales, supporting its applicability as a screening procedure for emotion regulation-related psychopathological disorders. To the best of our knowledge, EarlyScreen is the first work to use automatically extracted behavioral features to characterize both neural activity and the diagnostic status of emotion regulation-related disorders in young children. We present insights from mental health professionals supporting the utility of EarlyScreen and discuss considerations for its subsequent deployment.  more » « less
Award ID(s):
1839999 1815347
Author(s) / Creator(s):
; ; ; ; ;
Date Published:
Journal Name:
Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies
Page Range / eLocation ID:
1 to 39
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. Affective neuroscience research suggests that maturational changes in reward circuitry during adolescence present opportunities for new learning, but likely also contribute to increases in vulnerability for psychiatric disorders such as depression and substance abuse. Basic research in animal models and human neuroimaging has made progress in understanding the normal development of reward circuitry in adolescence, yet, few functional neuroimaging studies have examined puberty-related influences on the functioning of this circuitry. The goal of this study was to address this gap by examining the extent to which striatal activation and cortico-striatal functional connectivity to cues predicting upcoming rewards would be positively associated with pubertal status and levels of pubertal hormones (dehydroepiandrosterone, testosterone, estradiol). Participants included 79 adolescents (10-13 year olds; 47 girls) varying in pubertal status who performed a novel reward cue processing task during fMRI. Pubertal maturation was assessed using sex-specific standardized composite measures based on Tanner staging (self-report and clinical assessment) and scores from the Pubertal Development Scale. These composite measures were computed to index overall pubertal maturation as well as maturation of the adrenal and gonadal axes separately for boys and girls. Basal levels of circulating pubertal hormones were measured using immunoassays from three samples collected weekly upon awakening across a three-week period. Results indicated greater striatal activation and functional connectivity between nucleus accumbens (NAcc) and medial prefrontal cortex (mPFC) to reward cue (vs. no reward cue) on this task. Also, girls with higher levels of estradiol showed reduced activation in left and right caudate and greater NAcc-putamen connectivity. Girls with higher levels of testosterone showed greater NAcc connectivity with the anterior cingulate cortex and the insula. There were no significant associations in boys. Findings suggest that patterns of activation and connectivity in cortico-striatal regions are associated with reward cue processing, particularly in girls. Longitudinal follow-up neuroimaging studies are needed to fully characterize puberty-specific effects on the development of these neural regions and how such changes may contribute to pathways of risk or resilience in adolescence. 
    more » « less
  2. Abstract

    Social cognition may facilitate fathers' sensitive caregiving behavior. We administered the Why‐How Task, an fMRI task that elicits theory of mind processing, to expectant fathers (n = 39) who also visited the laboratory during their partner's pregnancy and provided a plasma sample for oxytocin assay. Three months postpartum, fathers reported their beliefs about parenting. When rating “Why” an action was being performed versus “How” the action was being performed (Why > How contrast), participants showed activation in regions theorized to support theory of mind, including the dorsomedial prefrontal cortex and superior temporal sulcus. Fathers' prenatal oxytocin levels predicted greater signal change during the Why > How contrast in the inferior parietal lobule. Both prenatal oxytocin and attunement parenting beliefs were associated with Why > How activation in the dorsolateral prefrontal cortex, a theory of mind region implicated in emotion regulation. Posterior parahippocampal gyrus and dorsolateral prefrontal cortex activation during the Why > How contrast predicted fathers' attunement parenting beliefs. In conclusion, fathers' neural activation when engaging in a theory of mind task was associated with their prenatal oxytocin levels and their postpartum attunement parenting beliefs. Results suggest biological and cognitive components of fathering may track with the theory of mind processing.

    more » « less
  3. Background

    Dysregulation of the corticotropin‐releasing factor (CRF) system has been observed in rodent models of binge drinking, with a large focus onCRFreceptor 1 (CRF‐R1). The role ofCRF‐binding protein (CRFBP), a key regulator ofCRFactivity, in binge drinking is less well understood. In humans, single‐nucleotide polymorphisms inCRHBPare associated with alcohol use disorder and stress‐induced alcohol craving, suggesting a role forCRFBPin vulnerability to alcohol addiction.


    The role and regulation ofCRFBPin binge drinking were examined in mice exposed to the drinking in the dark (DID) paradigm. Using in situ hybridization, the regulation ofCRFBP,CRF‐R1, andCRFmRNAexpression was determined in the stress and reward systems of C57BL/6J mice after repeated cycles ofDID. To determine the functional role ofCRFBPin binge drinking,CRFBPknockout (CRFBP KO) mice were exposed to 6 cycles ofDID, during which alcohol consumption was measured and compared to wild‐type mice.


    CRFBPmRNAexpression was significantly decreased in the prelimbic (PL) and infralimbic medial prefrontal cortex (mPFC) of C57BL/6J mice after 3 cycles and in thePLmPFCafter 6 cycles ofDID. No significant changes inCRForCRF‐R1 mRNAlevels were observed in mPFC, ventral tegmental area, bed nucleus of the stria terminalis, or amygdala after 3 cycles ofDID.CRFBP KOmice do not show significant alterations in drinking compared to wild‐type mice across 6 cycles of DID.


    These results reveal that repeated cycles of binge drinking alterCRFBPmRNAexpression in the mPFC, a region responsible for executive function and regulation of emotion and behavior, including responses to stress. We observed a persistent decrease inCRFBPmRNAexpression in the mPFCafter 3 and 6DIDcycles, which may allow for increasedCRFsignaling atCRF‐R1 and contribute to excessive binge‐like ethanol consumption.

    more » « less
  4. Objective Sudden unexpected death in epilepsy (SUDEP) is the leading cause of epilepsy-related mortality. Although lots of effort has been made in identifying clinical risk factors for SUDEP in the literature, there are few validated methods to predict individual SUDEP risk. Prolonged postictal EEG suppression (PGES) is a potential SUDEP biomarker, but its occurrence is infrequent and requires epilepsy monitoring unit admission. We use machine learning methods to examine SUDEP risk using interictal EEG and ECG recordings from SUDEP cases and matched living epilepsy controls. Methods This multicenter, retrospective, cohort study examined interictal EEG and ECG recordings from 30 SUDEP cases and 58 age-matched living epilepsy patient controls. We trained machine learning models with interictal EEG and ECG features to predict the retrospective SUDEP risk for each patient. We assessed cross-validated classification accuracy and the area under the receiver operating characteristic (AUC) curve. Results The logistic regression (LR) classifier produced the overall best performance, outperforming the support vector machine (SVM), random forest (RF), and convolutional neural network (CNN). Among the 30 patients with SUDEP [14 females; mean age (SD), 31 (8.47) years] and 58 living epilepsy controls [26 females (43%); mean age (SD) 31 (8.5) years], the LR model achieved the median AUC of 0.77 [interquartile range (IQR), 0.73–0.80] in five-fold cross-validation using interictal alpha and low gamma power ratio of the EEG and heart rate variability (HRV) features extracted from the ECG. The LR model achieved the mean AUC of 0.79 in leave-one-center-out prediction. Conclusions Our results support that machine learning-driven models may quantify SUDEP risk for epilepsy patients, future refinements in our model may help predict individualized SUDEP risk and help clinicians correlate predictive scores with the clinical data. Low-cost and noninvasive interictal biomarkers of SUDEP risk may help clinicians to identify high-risk patients and initiate preventive strategies. 
    more » « less
  5. Kazuhiro Maeshima (Ed.)

    Supranucleosomal chromatin structure, including chromatin domain conformation, is involved in the regulation of gene expression and its dysregulation has been associated with carcinogenesis. Prior studies have shown that cells in the buccal mucosa carry a molecular signature of lung cancer among the cigarette-smoking population, the phenomenon known as field carcinogenesis or field of injury. Thus, we hypothesized that chromatin structural changes in buccal mucosa can be predictive of lung cancer. However, the small size of the chromatin chain (approximately 20 nm) folded into chromatin packing domains, themselves typically below 300 nm in diameter, preclude the detection of alterations in intradomain chromatin conformation using diffraction-limited optical microscopy. In this study, we developed an optical spectroscopic statistical nanosensing technique to detect chromatin packing domain changes in buccal mucosa as a lung cancer biomarker: chromatin-sensitive partial wave spectroscopic microscopy (csPWS). Artificial intelligence (AI) was applied to csPWS measurements of chromatin alterations to enhance diagnostic performance. Our AI-enhanced buccal csPWS nanocytology of 179 patients at two clinical sites distinguished Stage-I lung cancer versus cancer-free controls with an area under the ROC curve (AUC) of 0.92 ± 0.06 for Site 1 (in-state location) and 0.82 ± 0.11 for Site 2 (out-of-state location).

    more » « less