18F-fluorodeoxyglucose (FDG)-positron emission tomography (PET) reveals altered brain metabolism in individuals with mild cognitive impairment (MCI) and Alzheimer’s disease (AD). Some biomarkers derived from FDG-PET by computer-aided-diagnosis (CAD) technologies have been proved that they can accurately diagnosis normal control (NC), MCI, and AD. However, existing FDG-PET-based researches are still insufficient for the identification of early MCI (EMCI) and late MCI (LMCI). Compared with methods based other modalities, current methods with FDG-PET are also inadequate in using the inter-region-based features for the diagnosis of early AD. Moreover, considering the variability in different individuals, some hard samples which are very similar with both two classes limit the classification performance. To tackle these problems, in this paper, we propose a novel bilinear pooling and metric learning network (BMNet), which can extract the inter-region representation features and distinguish hard samples by constructing the embedding space. To validate the proposed method, we collect 898 FDG-PET images from Alzheimer’s disease neuroimaging initiative (ADNI) including 263 normal control (NC) patients, 290 EMCI patients, 147 LMCI patients, and 198 AD patients. Following the common preprocessing steps, 90 features are extracted from each FDG-PET image according to the automatic anatomical landmark (AAL) template and then sent into the proposed network. Extensive fivefold cross-validation experiments are performed for multiple two-class classifications. Experiments show that most metrics are improved after adding the bilinear pooling module and metric losses to the Baseline model respectively. Specifically, in the classification task between EMCI and LMCI, the specificity improves 6.38% after adding the triple metric loss, and the negative predictive value (NPV) improves 3.45% after using the bilinear pooling module. In addition, the accuracy of classification between EMCI and LMCI achieves 79.64% using imbalanced FDG-PET images, which illustrates that the proposed method yields a state-of-the-art result of the classification accuracy between EMCI and LMCI based on PET images.

ObjectivesTo evaluate the relationship between systemic family dynamics and adolescent depression.MethodsAn offline survey was distributed to 4,109 students in grades 6–12, with the final analysis including 3,014 students (1,524 boys and 1,490 girls) aged 10–18 years. The questionnaire included the Self-Rating Scale of Systemic Family Dynamics (SSFD), the Self-Rating Depression Scale (SDS), and demographic characteristics.ResultsFamily dynamics were negatively correlated with depressive symptoms, with better family dynamics (high scores) associated with lower levels of depression based on the SDS score. After adjusting for sociodemographic characteristics, an ordinal multiclass logistic regression analysis identified family atmosphere (OR = 0.952, 95% CI: 0.948–0.956, p < 0.001) as the most important protective family dynamic against depression, followed by individuality (OR = 0.964, 95% CI: 0.960–0.968, p < 0.001). Latent class analysis (LCA) created the low family dynamic and high family dynamic groups. There were significant differences in the mean SDS scores between the two groups (45.52 ± 10.57 vs. 53.78 ± 11.88; p < 0.001) that persisted after propensity matching. Family atmosphere and individuation had a favorable diagnostic value for depression, with AUCs of 0.778 (95% CI: 0.760–0.796) and 0.710 (95% CI: 0.690–0.730), respectively. The diagnostic models for depression performed well.ConclusionPoor family dynamics may be responsible for adolescent depression. A variety of early intervention strategies focused on the family may potentially avoid adolescent depression.

No abstract