Background: Recent imaging studies of large datasets suggested that psychiatric disorders have common biological substrates. This study aimed to identify all the common neural substrates with connectomic abnormalities across four major psychiatric disorders by using the data-driven connectome-wide association method of multivariate distance matrix regression (MDMR). Methods: This study analyzed a resting functional magnetic resonance imaging dataset of 100 patients with schizophrenia, 100 patients with bipolar I disorder, 100 patients with bipolar II disorder, 100 patients with major depressive disorder, and 100 healthy controls (HCs). We calculated a voxel-wise 4,330 × 4,330 matrix of whole-brain functional connectivity (FC) with 8-mm isotropic resolution for each participant and then performed MDMR to identify structures where the overall multivariate pattern of FC was significantly different between each patient group and the HC group. A conjunction analysis was performed to identify common neural regions with FC abnormalities across these four psychiatric disorders. Results: The conjunction of the MDMR maps revealed that the four groups of patients shared connectomic abnormalities in distributed cortical and subcortical structures, which included bilateral thalamus, cerebellum, frontal pole, supramarginal gyrus, postcentral gyrus, lingual gyrus, lateral occipital cortex, and parahippocampus. The follow-up analysis based on pair-wise FC of these regions demonstrated that these psychiatric disorders also shared similar patterns of FC abnormalities characterized by sensory/subcortical hyperconnectivity, association/subcortical hypoconnectivity, and sensory/association hyperconnectivity. Conclusions: These findings suggest that major psychiatric disorders share common connectomic abnormalities in distributed cortical and subcortical regions and provide crucial support for the common network hypothesis of major psychiatric disorders.

Diabetic macular edema (DME) is a highly common cause of vision loss in patients with diabetes. Optical coherence tomography (OCT) is crucial in classifying DME and tracking the results of DME treatment. The presence of intraretinal cystoid fluid (IRC) and subretinal fluid (SRF) and the disruption of the ellipsoid zone (EZ), which is part of the photoreceptor layer, are three crucial factors affecting the best corrected visual acuity (BCVA). However, the manual segmentation of retinal fluid and the EZ from retinal OCT images is laborious and time-consuming. Current methods focus only on the segmentation of retinal features, lacking a correlation with visual acuity. Therefore, we proposed a modified U-net, a deep learning algorithm, to segment these features from OCT images of patients with DME. We also correlated these features with visual acuity. The IRC, SRF, and EZ of the OCT retinal images were manually labeled and checked by doctors. We trained the modified U-net model on these labeled images. Our model achieved Sørensen–Dice coefficients of 0.80 and 0.89 for IRC and SRF, respectively. The area under the receiver operating characteristic curve (ROC) for EZ disruption was 0.88. Linear regression indicated that EZ disruption was the factor most strongly correlated with BCVA. This finding agrees with that of previous studies on OCT images. Thus, we demonstrate that our segmentation network can be feasibly applied to OCT image segmentation and assist physicians in assessing the severity of the disease.

Background: Optical coherence tomography (OCT) is considered as a sensitive and noninvasive tool to evaluate the macular lesions. In patients with diabetes mellitus (DM), the existence of diabetic macular edema (DME) can cause significant vision impairment and further intravitreal injection (IVI) of anti–vascular endothelial growth factor (VEGF) is needed. However, the increasing number of DM patients makes it a big burden for clinicians to manually determine whether DME exists in the OCT images. The artificial intelligence (AI) now enormously applied to many medical territories may help reduce the burden on clinicians. Methods: We selected DME patients receiving IVI of anti-VEGF or corticosteroid at Taipei Veterans General Hospital in 2017. All macular cross-sectional scan OCT images were collected retrospectively from the eyes of these patients from January 2008 to July 2018. We further established AI models based on convolutional neural network architecture to determine whether the DM patients have DME by OCT images. Results: Based on the convolutional neural networks, InceptionV3 and VGG16, our AI system achieved a high DME diagnostic accuracy of 93.09% and 92.82%, respectively. The sensitivity of the VGG16 and InceptionV3 models was 96.48% and 95.15%., respectively. The specificity was corresponding to 86.67% and 89.63% for VGG16 and InceptionV3, respectively. We further developed an OCT-driven platform based on these AI models. Conclusion: We successfully set up AI models to provide an accurate diagnosis of DME by OCT images. These models may assist clinicians in screening DME in DM patients in the future.

The amyloid framework forms the central medical theory related to Alzheimer disease (AD), and the in vivo demonstration of amyloid positivity is essential for diagnosing AD. On the basis of a longitudinal cohort design, the study investigated clinical progressive patterns by obtaining cognitive and structural measurements from a group of patients with amnestic mild cognitive impairment (MCI); the measurements were classified by the positivity (Aβ+) or absence (Aβ−) of the amyloid biomarker. We enrolled 185 patients (64 controls, 121 patients with MCI). The patients with MCI were classified into two groups on the basis of their [¹⁸F]flubetaben or [¹⁸F]florbetapir amyloid positron-emission tomography scan (Aβ+ vs. Aβ−, 67 vs. 54 patients) results. Data from annual cognitive measurements and three-dimensional T1 magnetic resonance imaging scans were used for between-group comparisons. To obtain longitudinal cognitive test scores, generalized estimating equations were applied. A linear mixed effects model was used to compare the time effect of cortical thickness degeneration. The cognitive decline trajectory of the Aβ+ group was obvious, whereas the Aβ− and control groups did not exhibit a noticeable decline over time. The group effects of cortical thickness indicated decreased entorhinal cortex in the Aβ+ group and supramarginal gyrus in the Aβ− group. The topology of neurodegeneration in the Aβ− group was emphasized in posterior cortical regions. A comparison of the changes in the Aβ+ and Aβ− groups over time revealed a higher rate of cortical thickness decline in the Aβ+ group than in the Aβ− group in the default mode network. The Aβ+ and Aβ− groups experienced different APOE ε4 effects. For cortical–cognitive correlations, the regions associated with cognitive decline in the Aβ+ group were mainly localized in the perisylvian and anterior cingulate regions. By contrast, the degenerative topography of Aβ− MCI was scattered. The memory learning curves, cognitive decline patterns, and cortical degeneration topographies of the two MCI groups were revealed to be different, suggesting a difference in pathophysiology. Longitudinal analysis may help to differentiate between these two MCI groups if biomarker access is unavailable in clinical settings.