Predicting suicidality in late-life depression by 3D convolutional neural network and cross-sample entropy analysis of resting-state fMRI.

BACKGROUND: Predicting suicide is a pressing issue among older adults; however, predicting its risk is difficult. Capitalizing on the recent development of machine learning, considerable progress has been made in predicting complex behavior such as suicide. As depression remained the strongest risk for suicide, we aimed to apply deep learning algorithms to identify suicidality in a group with late-life depression (LLD). METHODS: We enrolled 83 patients with LLD, 35 of which were non-suicidal and 48 were suicidal, including 26 with only suicidal ideation and 22 with past suicide attempts, for resting-state functional magnetic resonance imaging (MRI). Cross-sample entropy (CSE) analysis was conducted to examine the complexity of MRI signals among brain regions. Three-dimensional (3D) convolutional neural networks (CNNs) were used, and the classification accuracy in each brain region was averaged to predict suicidality after sixfold cross-validation. RESULTS: We found brain regions with a mean accuracy above 75% to predict suicidality located mostly in default mode, fronto-parietal, and cingulo-opercular resting-state networks. The models with right amygdala and left caudate provided the most reliable accuracy in all cross-validation folds, indicating their neurobiological importance in late-life suicide. CONCLUSION: Combining CSE analysis and the 3D CNN, several brain regions were found to be associated with suicidality.

Mental Health of Children With Special Educational Needs and the Return to In-Person Learning After the COVID-19 Pandemic.

Importance: There is limited evidence on whether the quality of life and behavior of children with special educational needs (SEN) have improved or worsened since schools reopened after COVID-19-related school closures. Objective: To describe the changes in the mental well-being of children and adolescents with SEN during the initial 6 months of resuming in-person learning after COVID-19-related school closures. Design, Setting, and Participants: This repeated cross-sectional study reported data from surveys completed by parents and caregivers of children and adolescents aged 3 to 18 years with SEN studying at special schools in Hong Kong. The first cohort was obtained during COVID-19-related school closure in April 2020 (wave 1) and the second cohort was obtained 6 months after school resumption with data collection between July and October 2021 (wave 2). Data analysis occurred from January to June 2022. Exposure: Diagnosis of a disability or disorder that required school-based special educational programming. Main Outcomes and Measures: Children's emotional and behavioral difficulties (measured with the Strengths and Difficulties Questionnaire [SDQ]), quality of life (measured with the Pediatric Quality of Life Inventory [PedsQL]), lifestyle habits, parental stress, and parental well-being (measured with the PedsQL Family Impact Module) were assessed. Cross-sectional comparisons of well-being between the 2 waves were conducted using analysis of covariance, and multiple regression analysis was performed to identify factors associated with mental health outcomes in wave 2. Results: In wave 1, a total of 456 parents and caregivers of children with SEN (mean [SD] age, 7.44 [3.98] years; 315 boys [69.1%]; 141 girls [30.9%]) responded to the surveys. In wave 2, 519 parents and caregivers of children with SEN (mean [SD] age, 8.16 [4.47] years; 365 boys [70.3%]; 154 girls [29.7%]) responded. After school resumption, preschoolers aged 3 to 5 years with SEN had significantly fewer emotional difficulties (mean [SD] SDQ score, 3.26 [2.39] vs 2.68 [2.03]; standardized mean difference [SMD] = 0.26; 95% CI, 0.07-0.46; Bonferroni-corrected P = .04) and conduct difficulties (mean [SD] SDQ score, 2.88 [1.89] vs 2.41 [1.91]; SMD = 0.25; 95% CI, 0.05-0.44; Bonferroni-corrected P = .01), whereas adolescents had more conduct difficulties (mean [SD] SDQ score, 1.62 [1.50] vs 2.37 [3.02]; SMD = 0.41; 95% CI, 0.13-0.70; Bonferroni-corrected P = .049). The overall quality of life of school-aged children with SEN aged 6 to 11 years worsened after school resumption (mean [SD] PedsQL score, 67.52 [17.45] vs 60.57 [16.52]; SMD = 0.41; 95% CI, 0.19-0.62; Bonferroni-corrected P = .002). Conclusions and Relevance: The findings of this repeated cross-sectional study suggest that preschoolers with SEN had improved emotional and behavioral functioning when school resumed after COVID-19-related closures. School-aged children with SEN, adolescents with SEN, and children with intellectual disabilities were at risk of reduced quality of life, indicating that additional support should be offered to vulnerable groups as they return to schools.

A Pilot Study of Short-Course Oral Vitamin A and Aerosolised Diffuser Olfactory Training for the Treatment of Smell Loss in Long COVID.

Background: Olfactory dysfunction (OD) is a common neurosensory manifestation in long COVID. An effective and safe treatment against COVID-19-related OD is needed. Methods: This pilot trial recruited long COVID patients with persistent OD. Participants were randomly assigned to receive short-course (14 days) oral vitamin A (VitA; 25,000 IU per day) and aerosolised diffuser olfactory training (OT) thrice daily (combination), OT alone (standard care), or observation (control) for 4 weeks. The primary outcome was differences in olfactory function by butanol threshold tests (BTT) between baseline and end-of-treatment. Secondary outcomes included smell identification tests (SIT), structural MRI brain, and serial seed-based functional connectivity (FC) analyses in the olfactory cortical network by resting-state functional MRI (rs-fMRI). Results: A total of 24 participants were randomly assigned to receive either combination treatment (n = 10), standard care (n = 9), or control (n = 5). Median OD duration was 157 days (IQR 127-175). Mean baseline BTT score was 2.3 (SD 1.1). At end-of-treatment, mean BTT scores were significantly higher for the combination group than control (p < 0.001, MD = 4.4, 95% CI 1.7 to 7.2) and standard care (p = 0.009) groups. Interval SIT scores increased significantly (p = 0.009) in the combination group. rs-fMRI showed significantly higher FC in the combination group when compared to other groups. At end-of-treatment, positive correlations were found in the increased FC at left inferior frontal gyrus and clinically significant improvements in measured BTT (r = 0.858, p < 0.001) and SIT (r = 0.548, p = 0.042) scores for the combination group. Conclusions: Short-course oral VitA and aerosolised diffuser OT was effective as a combination treatment for persistent OD in long COVID.

Internet-based and mobile-based cognitive behavioral therapy for chronic diseases: a systematic review and meta-analysis.

Positive adjustment to chronic diseases reduces psychiatric comorbidity and enhances quality of life. Very little is known about the benefit of internet-based and mobile-based Cognitive Behavioral Therapy (IM-CBT) on physical outcomes and its reciprocal interactions with psychiatric outcomes, the active therapeutic elements, and effect moderators among people with major chronic medical conditions. In this systematic review and meta-analysis (PROSPERO: CRD42022265738), CINAHL of Systematic Reviews, MEDLINE, PsycINFO, PubMed, Web of Science are systematically searched up to 1 June 2022, for randomized controlled trials (RCTs) comparing IM-CBT against non-CBT control condition(s) among people with chronic disease(s). Primary outcomes include improvements in psychiatric symptoms (depressive, anxiety, PTSD symptoms, general psychological distress) from baseline to post-intervention and follow-ups. Secondary outcomes include improvements in physical distress (physical symptoms, functional impairment, self-rated ill health, objective physiological dysfunction). Among 44 RCTs (5077 patients with seven different chronic diseases), IM-CBT improves depressive symptoms, anxiety symptoms, and general psychological distress at post-intervention and across follow-ups, and improves physical distress and functional impairment at post-intervention. Preliminary evidence suggests that behavioral modification and problem-solving could be necessary components to reduce psychiatric symptoms in IM-CBT, whereas cognitive restructuring, psychoeducation, and mindfulness elements relate to reduced physical distress. IM-CBT shows stronger benefits in chronic pain, cancer, arthritis, and cardiovascular disease, relative to other conditions. Changes in psychiatric symptoms and physical distress prospectively predict each other over time. IM-CBT is an effective intervention for comprehensive symptom management among people with chronic diseases.

Automatic diagnosis of late-life depression by 3D convolutional neural networks and cross-sample Entropy analysis from resting-state fMRI.

Resting-state fMRI has been widely used in investigating the pathophysiology of late-life depression (LLD). Unlike the conventional linear approach, cross-sample entropy (CSE) analysis shows the nonlinear property in fMRI signals between brain regions. Moreover, recent advances in deep learning, such as convolutional neural networks (CNNs), provide a timely application for understanding LLD. Accurate and prompt diagnosis is essential in LLD; hence, this study aimed to combine CNN and CSE analysis to discriminate LLD patients and non-depressed comparison older adults based on brain resting-state fMRI signals. Seventy-seven older adults, including 49 patients and 28 comparison older adults, were included for fMRI scans. Three-dimensional CSEs with volumes corresponding to 90 seed regions of interest of each participant were developed and fed into models for disease classification and depression severity prediction. We obtained a diagnostic accuracy > 85% in the superior frontal gyrus (left dorsolateral and right orbital parts), left insula, and right middle occipital gyrus. With a mean root-mean-square error (RMSE) of 2.41, three separate models were required to predict depressive symptoms in the severe, moderate, and mild depression groups. The CSE volumes in the left inferior parietal lobule, left parahippocampal gyrus, and left postcentral gyrus performed best in each respective model. Combined complexity analysis and deep learning algorithms can classify patients with LLD from comparison older adults and predict symptom severity based on fMRI data. Such application can be utilized in precision medicine for disease detection and symptom monitoring in LLD.

Preterm postnatal complications and risk of attention-deficit/hyperactivity disorder.

AIM: To investigate the association between the risk of attention-deficit/hyperactivity disorder (ADHD) and preterm birth and determine how postnatal complications in children born preterm is associated with the risk of ADHD. METHOD: This population-based cohort study used data from the Hong Kong electronic medical records. We followed 359 614 children (48% female; 6-17 years old, mean 11 years 7 months, SD 3 years 2 months) born in public hospitals in Hong Kong from 1st January 2004 to 31st December 2014 and collected medical records and demographic details for mothers and children until 11th November 2020. RESULTS: The risk of ADHD was 4.0% in children born at term and 5.1% in children born preterm. The odds ratio for ADHD was 2.08 (95% confidence interval [CI] 1.64-2.64) for children born extremely preterm, 1.64 (95% CI 1.46-1.85) for children born very preterm, and 1.15 (95% CI 1.08-1.23) for children born late preterm. Among preterm postnatal complications, only early respiratory disease, retinopathy of prematurity (ROP), and intraventricular haemorrhage were significant predictors of ADHD after controlling for preterm birth, other risk factors, and sociodemographic variables. The excess risk of ADHD among children born very preterm or late preterm could be partly explained by respiratory disease. ROP partially mediated the risk of ADHD in children born very preterm. INTERPRETATION: Children born preterm in all subcategories, from extremely preterm to late preterm, have increased risk of ADHD. Early respiratory infection partially mediates the risk of ADHD in children born preterm.

Relationships between Inflammation and Age-Related Neurocognitive Changes.

The relationship between inflammation and age-related neurocognitive changes is significant, which may relate to the age-related immune dysfunctions characterized by the senescence of immune cells and elevated inflammatory markers in the peripheral circulation and the central nervous system. In this review, we discuss the potential mechanisms, including the development of vascular inflammation, neuroinflammation, organelle dysfunctions, abnormal cholesterol metabolism, and glymphatic dysfunctions as well as the role that the key molecules play in the immune-cognition interplay. We propose potential therapeutic pharmacological and behavioral strategies for ameliorating age-related neurocognitive changes associated with inflammation. Further research to decipher the multidimensional roles of chronic inflammation in normal and pathological aging processes will help unfold the pathophysiological mechanisms underpinning neurocognitive disorders. The insight gained will lay the path for developing cost-effective preventative measures and the buffering or delaying of age-related neurocognitive decline.

Low assets predict persistent depression through living difficulties amid large-scale disasters: A cohort study.

BACKGROUND: In face of large-scale disasters, persons with fewer assets are at greater risk of persistent poorer mental health than persons with more assets. Everyday daily routine disruptions and financial hardship could mediate this association. METHODS: This prospective population-representative study in Hong Kong aimed to investigate the relation between assets during the acute phase of COVID-19 (February-August 2020, T1) and persistent probable depression from T1 to March-August 2021 (T2), as well as the mediating effects of daily routine disruptions and financial hardship on the assets-depression association. RESULTS: Low assets at T1 prospectively related to persistent probable depression from T1 to T2. Primary routine disruptions (i.e., healthy eating and sleep) at T1 and financial hardship at T2 were found to fully mediate the association between T1 assets and persistent probable depression. LIMITATIONS: Persistent probable depression reported on the PHQ-9 should be further verified with clinical diagnoses/interviews. CONCLUSIONS: The COVID-19 pandemic was accompanied by a global economic downturn. Persons who have fewer assets could be at greater risk of depression during this period. Our findings suggest a need to provide behavioral and financial assistance to persons with fewer assets in the short run and a need to ensure that everyone has adequate assets to mitigate the mental health consequences of the COVID-19 pandemic in the long run.

Neurobehavioral Impairment in Pediatric Brain Tumor Survivors: A Meta-Analysis.

Purpose: The neurocognitive outcomes of pediatric brain tumor survivors have been extensively studied but the risk and predictors for neurobehavioral impairment are less clearly defined. We systematically analyzed the rates of emotional, psychosocial, and attention problems in pediatric brain tumor survivors. Methods: PubMed, Web of Science, Embase, Scopus, and Cochrane were searched for articles published between January 2012 to April 2022. Eligible studies reported neurobehavioral outcomes for PBTS aged 2 to <23 years with a brain tumor diagnosis before 18 years of age. A random-effect meta-analysis was performed in R. Results: The search yielded 1187 unique publications, of which 50 were included in the quantitative analysis. The estimated risk of having emotional, psychosocial, and attention problems were 15% (95%CI 10−20%), 12% (95%CI 9−16%), and 12% (95%CI 9−16%), respectively. PBTS were more likely to have emotional difficulties (Hedge's g = 0.43 [95%CI 0.34−0.52]), psychosocial problems (Hedge's g = 0.46 [95%CI 0.33−0.58]), and attention problems (Hedge's g = 0.48 [95%CI 0.34−0.63]) compared to normal/healthy control subjects. There was no significant difference in the rates of neurobehavioral impairment between children with and without history of cranial radiotherapy. Conclusions: PBTS are at elevated risk of neurobehavioral impairment. Neurobehavioral monitoring should be considered as the standard of care for PBTS.

Disrupted daily routines mediate the socioeconomic gradient of depression amid public health crises: A repeated cross-sectional study.

OBJECTIVE: There is a socioeconomic gradient to depression risks, with more pronounced inequality amid macroenvironmental potential traumatic events. Between mid-2019 and mid-2020, the Hong Kong population experienced drastic societal changes, including the escalating civil unrest and the COVID-19 pandemic. We examined the change of the socioeconomic gradient in depression and the potential intermediary role of daily routine disruptions. METHOD: We conducted repeated territory-wide telephone surveys in July 2019 and July 2020 with 1112 and 2034 population-representative Cantonese-speaking Hong Kong citizens above 15 years old, respectively. Stratified by year, we examined the association between socioeconomic indicators (education attainment, household income, employment status and marital status) and probable depression (nine-item Patient Health Questionnaire [PHQ-9] ⩾ 10) using logistic regression. Differences in the socioeconomic gradient between 2019 and 2020 were tested. Finally, we performed a path analysis to test for the mediating role of daily routine disruptions. RESULTS: Logistic regression showed that higher education attainment in 2019 and being married in 2020 were protective against probable depression. Interaction analysis showed that the inverse association of higher education attainment with probable depression attenuated in 2020 but that of being married increased. Path analysis showed that the mediated effects through daily routine disruptions accounted for 95.9% of the socioeconomic gradient of probable depression in 2020, compared with 13.1% in 2019. CONCLUSION: From July 2019 to July 2020, the mediating role of daily routine disruptions in the socioeconomic gradient of depression in Hong Kong increased. It is thus implied that infection control measures should consider the relevant potential mental health impacts accordingly.

Trends of depression and anxiety during massive civil unrest and COVID-19 in Hong Kong, 2019-2020.

Both COVID-19 and unrest are posing a significant threat to population mental health across the globe. This study examined trends of probable depression and anxiety during a time of civil unrest and concurrent COVID-19 in Hong Kong. Four random digit dialing telephone surveys were conducted in July 2019 (n = 1112), February-March 2020 (n = 2003), April-May 2020 (n = 2008), and July-August 2020 (n = 2034). The prevalence of probable depression increased from 25.7% (95% CI: 23.2-28.3) in July 2019 to 28.2% (95% CI: 26.2-30.1) in February-March 2020, and then decreased to 15.3% (95% CI: 14.0-17.0) in April-May 2020 and 13.7% (95% CI: 12.2-15.2) in July-August 2020. The prevalence of probable anxiety was 19.2% (95% CI: 17.5-20.9) in February-March 2020 and then stabilized in April-May 2020 and July-August 2020 (14.1%, 95% CI: 12.0-15.8). Probable depression and anxiety were more prevalent among persons with high relative to low daily routine disruptions. Combined high unrest-COVID-19 stress was associated with probable depression and anxiety across all persons; high unrest stress alone was associated with probable mental disorders at high daily routine disruptions. Civil unrest and COVID-19 are jointly associated with depression and anxiety among Hong Kong citizens. While population mental health improved, daily routine disruptions is a risk factor of mental disorders at every time-point.

Brain Microstructural Changes Associated With Neurocognitive Outcome in Intracranial Germ Cell Tumor Survivors.

BACKGROUND: Childhood intracranial germ cell tumor (GCT) survivors are prone to radiotherapy-related neurotoxicity, which can lead to neurocognitive dysfunctions. Diffusion kurtosis imaging (DKI) is a diffusion MRI technique that is sensitive to brain microstructural changes. This study aimed to investigate the association between DKI metrics versus cognitive and functional outcomes of childhood intracranial GCT survivors. METHODS: DKI was performed on childhood intracranial GCT survivors (n = 20) who had received cranial radiotherapy, and age and gender-matched healthy control subjects (n = 14). Neurocognitive assessment was performed using the Hong Kong Wechsler Intelligence Scales, and functional assessment was performed using the Lansky/Karnofsky performance scales (KPS). Survivors and healthy controls were compared using mixed effects model. Multiple regression analyses were performed to determine the effects of microstructural brain changes of the whole brain as well as the association between IQ and Karnofsky scores and the thereof. RESULTS: The mean Intelligence Quotient (IQ) of GCT survivors was 91.7 (95% CI 84.5 - 98.8), which was below the age-specific normative expected mean IQ (P = 0.013). The mean KPS score of GCT survivors was 85.5, which was significantly lower than that of controls (P < 0.001). Cognitive impairments were significantly associated with the presence of microstructural changes in white and grey matter, whereas functional impairments were mostly associated with microstructural changes in white matter. There were significant correlations between IQ versus the mean diffusivity (MD) and mean kurtosis (MK) of specific white matter regions. The IQ scores were negatively correlated with the MD of extensive grey matter regions. CONCLUSION: Our study identified vulnerable brain regions whose microstructural changes in white and grey matter were significantly associated with impaired cognitive and physical functioning in survivors of pediatric intracranial GCT.

Greater white matter hyperintensities and the association with executive function in suicide attempters with late-life depression.

Late-life depression (LLD) is associated with greater risk of suicide and white matter hyperintensities (WMH), which are also found in suicide attempters regardless of age. Greater periventricular WMH are related to worse cognitive function. We investigated the spatial distribution of WMH in suicide attempters with LLD and its association with cognitive function. We recruited 114 participants with LLD (34 with history of suicide attempt and 80 without) and 47 older adult controls (individuals without LLD or history of suicide attempt). WMH were quantified by an automated segmentation algorithm and were classified into different regions. Suicide attempters with LLD had significantly higher global WMH (F3, 150 = 2.856, p = 0.039) and periventricular WMH (F3, 150 = 3.635, p = 0.014) compared to other groups. Suicide attempters with high WMH had significantly lower executive function, which could be an underlying mechanism for cognitive decline in older adults with suicidality.

Psychiatric symptoms and behavioral adjustment during the COVID-19 pandemic: evidence from two population-representative cohorts.

This study examined prevalences of anxiety and depression and their correlations with daily routines among Hong Kong Chinese during the COVID-19 pandemic. Random digit dialing recruited two population-representative samples of 6029 residents during a period of low infection and limited intervention (survey 1: n = 4021) and high incidence and intensive measures (survey 2: n = 2008). Prevalence of anxiety for survey 1 and survey 2 were 14.9% and 14% and depression were 19.6% and 15.3%, respectively. Increased odds of anxiety and depression were associated with disrupted routines and lower socioeconomic status in both surveys, whereas depression was inversely related to the novel preventive routine of avoiding going to crowded places in survey 1. The prevalences of anxiety and depression were higher than preceding public health/social crises. A heavier burden of psychiatric conditions was evidenced amongst people experiencing disrupted daily routines across different phases of the pandemic and without novel preventive routines in the early phase.

Civil unrest, COVID-19 stressors, anxiety, and depression in the acute phase of the pandemic: a population-based study in Hong Kong.

PURPOSE: To examine the joint associations of civil unrest and COVID-19 with probable anxiety and depression during the first half of 2020 in Hong Kong. Associations were compared between persons with low or high assets. METHODS: A population-representative sample of 4011 Hong Kong Chinese residents aged 15 years or older were recruited between February and May 2020. Respondents reported current anxiety and depressive symptoms, unrest stress, COVID-19 stress, assets (savings and home ownership), and demographics. RESULTS: Stress due to unrest and COVID-19 was associated with higher prevalence of probable anxiety and depression; persons with both stressors had higher prevalence. This pattern was consistent among persons with low or high assets, but the probabilities of mental disorder were substantially higher among persons with fewer assets. CONCLUSIONS: The effect of stressors on probable anxiety and depression are cumulative: persons with stress due to civil unrest and to COVID-19 reported more mental disorders than persons with stress due to only one, or none of these factors. Overall high assets appear to buffer the consequences of stressors, lowering the risk of mental disorder.

Probable anxiety and components of psychological resilience amid COVID-19: A population-based study.

BACKGROUND: This study examined the associations between components of psychological resilience and mental health at different levels of exposure to COVID-19 stressors. METHODS: A population-representative sample of 4,021 respondents were recruited and assessed between February 25th and March 19th, 2020. Respondents reported current anxiety symptoms (7-item Generalized Anxiety Disorder scale [GAD-7]), cognitive components (perceived ability to adapt to change, tendency to bounce back after adversities) and behavioral components (regularity of primary and secondary daily routines) of resilience, worry about COVID-19 infection, and sociodemographics. RESULTS: Logistic regression revealed that cognitive and behavioral components of resilience were not correlated with probable anxiety (GAD-7≥10) among those reporting no worry. Among respondents who were worried, all resilient components were inversely associated with probable anxiety. Specifically, propensity to bounce back and regular primary routines were more strongly related to lower odds of probable anxiety among those reporting lower levels of worry. LIMITATIONS: The cross-sectional design limits causal inference. Second, other resilient components and some key daily routines that could be related to better mental health were not assessed. Third, generalizability of the findings to other similar major cities is uncertain because cases and deaths due to COVID-19 in Hong Kong have been comparatively lower. CONCLUSIONS: To foster mental health, cultivation of confidence in one's ability to adapt to change and a propensity to bounce back from hardship should be coupled with sustainment of regular daily routines. Such assessment and intervention protocols could be more relevant to those who suffer heightened levels of exposure to COVID-19 stressors.

Social Capital, Income Loss, and Psychobehavioral Responses amid COVID-19: A Population-Based Analysis.

This study examined the associations of perceived social capital and income change since the outbreak with probable depression and preventive behaviors during the COVID-19 pandemic in Hong Kong. Random digit dialing recruited a population-representative sample of 3011 Hong Kong Chinese aged ≥ 15 years (mean = 44, 55% females) between February 25 and April 29 2020. Respondents reported social capital (perceived interpersonal trust, social harmony, and sense of belonging), income change since the outbreak (loss vs. gain/no change), depressive symptoms, preventive behaviors, and demographics. Controlling for sociodemographics, lack of perceived interpersonal trust was associated with probable depression and avoiding contact with people with respiratory symptoms. Lack of perceived sense of belonging was associated with probable depression and decreased odds of adopting preventive behaviors. Lack of perceived social harmony was associated with probable depression and increased odds of used face masks among respondents with income loss only. Our results suggest that social capital is related to lower risk of depression and to higher chance of used face masks particularly among those experiencing income loss related to COVID-19. Prevention of mental health problems and promotion of effective preventive behaviors could be implemented by focusing on support for those who are socioeconomically disadvantaged.

Depression Scale Prediction with Cross-Sample Entropy and Deep Learning.

A two-stage deep learning-based scheme is presented to predict the Hamilton Depression Scale (HAM-D) in this study. First, the cross-sample entropy (CSE) that allows assessing the degree of similarity of two data series are evaluated for the 90 brain regions of interest partitioned according to Automated Anatomical Labeling. The obtained CSE maps are then converted to 3D CSE volumes to serve as the inputs to the deep learning network models for the HAM-D scale level classification and prediction. The efficacy of the proposed scheme was illustrated by the resting-state functional magnetic resonance imaging data from 38 patients. From the results, the root mean square errors for the HAM-D scale prediction obtained during training, validation, and testing were 2.73, 2.66, and 2.18, which were less than those of a scheme having only a regression stage.

Cognitive Reserve Moderates Effects of White Matter Hyperintensity on Depressive Symptoms and Cognitive Function in Late-Life Depression.

INTRODUCTION: White matter hyperintensity (WMH) has been regarded as one of the major contributor of the vascular hypothesis of late-life depression (LLD) and cognitive decline in the elderly. On the other hand, cognitive reserve (CR) has long been hypothesized to provide resilience and adaptability against age- and disease-related insults. This study examined the role of CR, using proxy of education, in moderating the association between WMH and clinical LLD expression. METHODS: A total of 54 elderly diagnosed with major depressive disorder and 38 matched healthy controls participated in this study. They received MRI scanning and a battery of neuropsychological tests. WMH was quantified by an automated segmentation algorithm. Linear regression analyses were conducted separately in the LLD and control groups to examine the effects of WMH, education and their interaction in depression severity and various cognitive domains. RESULTS: WMH was significantly and negatively associated with executive function only in the healthy controls. In patients with LLD, we observed a significant interactive effect in education on the association between WMH and depression severity and language domain (category fluency task). Specifically, those with high education showed less depressive symptoms and cognitive decline as WMH increased. CONCLUSION: WMH is associated with lower cognitive function. However, in patients with LLD, high education attenuates the deleterious effect of WMH on mood and cognition. Therefore, CR appears to exert a protective effect on neurocognitive functioning in people with LLD.