Search

Background: To investigate the influence of artificial intelligent (AI) based on deep learning on the diagnostic performance and consistency of inexperienced cardiovascular radiologists.Methods: We enrolled 196 patents who had undergone both CCTA and invasive coronary angiography (ICA) within 6 months. Four readers with less cardiovascular experience (Reader 1 to Reader 4) and two cardiovascular radiologists (level II, Reader 5 and Reader 6) evaluated all images for ≥50% coronary artery stenosis, with ICA as the gold standard. Reader 3 and Reader 4 interpreted with aid from an AI system, and the other four readers interpreted without the AI system. The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and accuracy (area under the receiver operating characteristic curve (AUC)) of six readers were calculated at the patient and vessel levels. Additionally, we evaluated interobserver consistency between Reader 1 and Reader 2, Reader 3 and Reader 4, and Reader 5 and Reader 6.Results: The AI system had 94% and 78% sensitivity at the patient and vessel levels, respectively, which were higher than Reader 5 and Reader 6. Reader 3 and Reader 4 aided by AI had a higher sensitivity (range: +7.2%~ +16.6% and +5.9%~ +16.1%, respectively) and NPV (range: +3.7%~ +13.4% and +2.7%~ +4.2%, respectively) than Reader 1 and Reader 2 without AI. There was good interobserver consistency between Reader 3 and Reader 4 in interpreting ≥50% stenosis (Kappa value= 0.75 and 0.80 at the patient and vessel levels, respectively). Only Reader 1 and Reader 2 had poor consistency (Kappa value= 0.25 and 0.37). Reader 5 and Reader 6 had moderate agreement (Kappa value= 0.55 and 0.61).Conclusions: Our study showed that using AI could effectively increase the sensitivity of inexperienced readers and significantly improve consistency in diagnosing coronary stenosis via CCTA.Trial registration: The clinical trial registration number: ChiCTR1900021867Name of registry: Diagnostic performance of artificial intelligence assisted coronary computed tomography angiography for the assessment of coronary atherosclerotic stenosis

Background To investigate the influence of artificial intelligence (AI) based on deep learning on the diagnostic performance and consistency of inexperienced cardiovascular radiologists. Methods We enrolled 196 patents who had undergone both coronary computed tomography angiography (CCTA) and invasive coronary angiography (ICA) within 6 months. Four readers with less cardiovascular experience (Reader 1–Reader 4) and two cardiovascular radiologists (level II, Reader 5 and Reader 6) evaluated all images for ≥ 50% coronary artery stenosis, with ICA as the gold standard. Reader 3 and Reader 4 interpreted with AI system assistance, and the other four readers interpreted without the AI system. The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and accuracy (area under the receiver operating characteristic curve (AUC)) of the six readers were calculated at the patient and vessel levels. Additionally, we evaluated the interobserver consistency between Reader 1 and Reader 2, Reader 3 and Reader 4, and Reader 5 and Reader 6. Results The AI system had 94% and 78% sensitivity at the patient and vessel levels, respectively, which were higher than that of Reader 5 and Reader 6. AI-assisted Reader 3 and Reader 4 had higher sensitivity (range + 7.2–+ 16.6% and + 5.9–+ 16.1%, respectively) and NPVs (range + 3.7–+ 13.4% and + 2.7–+ 4.2%, respectively) than Reader 1 and Reader 2 without AI. Good interobserver consistency was found between Reader 3 and Reader 4 in interpreting ≥ 50% stenosis (Kappa value = 0.75 and 0.80 at the patient and vessel levels, respectively). Only Reader 1 and Reader 2 showed poor interobserver consistency (Kappa value = 0.25 and 0.37). Reader 5 and Reader 6 showed moderate agreement (Kappa value = 0.55 and 0.61). Conclusions Our study showed that using AI could effectively increase the sensitivity of inexperienced readers and significantly improve the consistency of coronary stenosis diagnosis via CCTA. Trial registration Clinical trial registration number: ChiCTR1900021867. Name of registry: Diagnostic performance of artificial intelligence-assisted coronary computed tomography angiography for the assessment of coronary atherosclerotic stenosis.

Objective: This study aimed at exploring the impact of patient-related, vessel-related, image quality-related and cardiovascular risk factors on coronary computed tomographic angiography (CCTA) interpretability using 256-detector row computed tomography (CT). Methods: One hundred ten patients who underwent CCTA and Invasive Coronary Angiography (ICA) were consecutively, retrospectively enrolled from January 2018 to October 2018. Using ICA as the reference standard, ≥50% diameter stenosis was defined as the cut-off criterion to detect the diagnostic performance of CCTA. Diagnostic reproducibility was investigated by calculating the interrater reproducibility of CCTA. Multiple logistic regression models were performed to evaluate the impact of 14 objective factors. Results: A total of 1019 segments were evaluated. The per-segment sensitivity, specificity, accuracy, positive predictive value, and negative predictive value of CCTA were 76.8%, 93.7%, 91.2%, 67.8%, and 95.9%, respectively. The per-segment diagnostic reproducibility was 0.44 for CCTA. Regarding accuracy, a negative association was found for stenosis severity, calcium load, and hyperlipidaemia. Regarding sensitivity, calcium load and diabetes mellitus (DM) were positively related. Regarding specificity, a negative correlation was observed between stenosis severity and calcium load. Regarding interrater reproducibility, stenosis severity and calcium load were negatively associated, whereas male sex and the signal-to-noise ratio (SNR) were positively related (all p Conclusion: Per-segment 256-detector row CCTA performance was optimal in stenosis-free or occluded segments. Heavier calcium load was associated with poorer CCTA interpretability. On the one hand, our findings confirmed the rule-out value of CCTA; on the other hand, improvements in calcium subtractions and deep learning-based tools are suggested to improve CCTA diagnostic interpretability.

To investigate the accuracy, diagnostic confidence, and interobserver agreement of subtraction coronary CT angiography (CCTA) versus invasive coronary angiography on 320-row CT in coronary segments with severe or non-severe calcification. Sixty-four patients (33 men, 66.6 ± 8.2 years) with suspected coronary artery disease (CAD) were prospectively enrolled from October 2019 to June 2020. The cross-sectional circumferential extent of calcification was used to classify calcified segments as non-severely ( < 180°) or severely calcified ( ≥ 180°). Three independent, blinded radiologists evaluated the severity of coronary stenosis. Interobserver agreement was evaluated using Fleiss’ kappa (κ). A multiple-reader multiple-case receiver operating characteristic (ROC) method was conducted, and diagnostic accuracy was measured using the mean areas under the ROC curves (AUCs), with ≥ 50% stenosis as a cut-off. Diagnostic confidence, diagnostic accuracy, and interobserver agreement were compared between CCTA with or without subtraction information in severely and non-severely calcified segments. In cases with severe calcification (51 patients, 146 segments), CCTA with subtraction information achieved better diagnostic accuracy (per-patient AUC: 0.73 vs 0.57, p = 0.03; per-segment AUC: 0.85 vs 0.62, p = 0.01), diagnostic confidence (3.7 vs 2.6, p < 0.001), and interobserver agreement (κ: 0.59 vs 0.30). Diagnostic accuracy (per-patient AUC: 0.81 vs 0.93, p = 0.30; per-patient AUC: 0.79 vs 0.82, p = 0.54) was not increased in cases with non-severe calcification (13 patients, 190 segments). CCTA with subtraction information achieved better diagnostic accuracy in cases of severe calcification (circumferential extent ≥ 180°). However, for non-severe calcification (circumferential extent < 180°), the effect of calcium subtraction was unclear, as it did not improve diagnostic accuracy. • Subtraction coronary CT angiography achieves better diagnostic accuracy, higher diagnostic confidence, and increased interobserver agreement for severe calcification (circumferential extent ≥ 180°). • Calcium subtraction does not improve the diagnostic accuracy of coronary CT angiography for calcification with a circumferential extent of < 180°.

Although both the granular layer of the prefrontal cortex (PFC) and schizophrenia are unique in primates, especially humans, their linkage is unclear. Here, we tested whether schizophrenia is associated with expression profiles of the granule cell (GC)-related genes in the human PFC. We identified 14 candidate GC-related genes with gradually increased expression levels along the gradient of the agranular, dysgranular, light-granular, and granular prefrontal regions based on the densely sampled gene expression data of 6 postmortem human brains, and with more than 10-fold expression in neurons than other cell types based on the single-cell RNA-sequencing data of the human PFC. These GC-related genes were functionally associated with synaptic transmission and cell development and differentiation. The identified 14 GC-related genes were significantly enriched for schizophrenia, but not for depression and bipolar disorder. The expression levels of the 4 stable schizophrenia- and GC-related genes were spatially correlated with gray matter volume differences in the PFC between patients with schizophrenia and healthy controls. This study provides a set of candidate genes for the human prefrontal GCs and links expression profiles of the GC-related genes to the prefrontal structural impairments in schizophrenia.

Background and Purpose: The human supplementary motor area (SMA) contains two functional subregions of the SMA proper and preSMA; however, the reorganization patterns of the two SMA subregions after stroke remain uncertain. Meanwhile, a focal subcortical lesion may affect the overall functional reorganization of brain networks. We sought to identify the differential reorganization of the SMA subregions after subcortical stroke using the resting-state functional connectivity (rsFC) analysis.Methods: Resting-state functional MRI was conducted in 25 patients with chronic capsular stroke exhibiting well-recovered global motor function (Fugl–Meyer score >90). The SMA proper and preSMA were identified by the rsFC-based parcellation, and the rsFCs of each SMA subregion were compared between stroke patients and healthy controls.Results: Despite common rsFC with the fronto-insular cortex (FIC), the SMA proper and preSMA were mainly correlated with the sensorimotor areas and cognitive-related regions, respectively. In stroke patients, the SMA proper and preSMA exhibited completely different functional reorganization patterns: the former showed increased rsFCs with the primary sensorimotor area and caudal cingulate motor area (CMA) of the motor execution network, whereas the latter showed increased rsFC with the rostral CMA of the motor control network. Both of the two SMA subregions showed decreased rsFC with the FIC in stroke patients; the preSMA additionally showed decreased rsFC with the prefrontal cortex (PFC).Conclusion: Although both SMA subregions exhibit functional disconnection with the cognitive-related areas, the SMA proper is implicated in the functional reorganization within the motor execution network, whereas the preSMA is involved in the functional reorganization within the motor control network in stroke patients.

Background Respective changes in resting-state cerebral blood flow (CBF) and functional connectivity in schizophrenia have been reported. However, their coupling alterations in schizophrenia remain largely unknown. Methods 89 schizophrenia patients and 90 sex- and age-matched healthy controls underwent resting-state functional MRI to calculate functional connectivity strength (FCS) and arterial spin labeling imaging to compute CBF. The CBF-FCS coupling of the whole gray matter and the CBF/FCS ratio (the amount of blood supply per unit of connectivity strength) of each voxel were compared between the 2 groups. Results Whole gray matter CBF-FCS coupling was decreased in schizophrenia patients relative to healthy controls. In schizophrenia patients, the decreased CBF/FCS ratio was predominantly located in cognitive- and emotional-related brain regions, including the dorsolateral prefrontal cortex, insula, hippocampus and thalamus, whereas an increased CBF/FCS ratio was mainly identified in the sensorimotor regions, including the putamen, and sensorimotor, mid-cingulate and visual cortices. Conclusion These findings suggest that the neurovascular decoupling in the brain may be a possible neuropathological mechanism of schizophrenia.

BackgroundAs a disconnection syndrome, schizophrenia has shown impaired resting-state functional connectivity (rsFC) in the orbitofrontal cortex (OFC); however, the OFC is a rather heterogeneous region and the rsFC changes in the OFC subregions remain unknown.MethodA total of 98 schizophrenia patients and 102 healthy controls underwent resting-state functional MRI using a sensitivity-encoded spiral-in imaging sequence (SENSE-SPIRAL) to reduce susceptibility-induced signal loss and distortion. The OFC subregions were defined according to a previous parcellation study that divided the OFC into the anterior (OFCa), medial (OFCm), posterior (OFCp), intermediate (OFCi), and lateral (OFCl) subregions. The rsFC was compared using two-way repeated-measures ANOVA.ResultsWhether or not global signal regression, compared with healthy controls, schizophrenia patients consistently exhibited decreased rsFC between the left OFCi and the left middle temporal gyrus and the right middle frontal gyrus (MFG), between the right OFCi and the right MFG and the left inferior frontal gyrus, between the right OFCm and the middle cingulate cortex and the left Rolandic operculum. These rsFC changes still remained significant even after cortical atrophy correction.ConclusionsThese findings suggest a selective functional disconnection of the OFC subregions in schizophrenia, and provide more precise information about the functional disconnections of the OFC in this disorder.

This paper presents an improved SURF algorithm based on gradient and amplitude pre-computation. This method is based on the traditional SURF algorithm and replaces the complex Gauss calculation with integral graph search. There are extreme value features and high frequency noise describing the edge and sharpness of the image. In order to measure the relationship between them, it is necessary to introduce the signal-to-noise ratio measurement index to highlight the effective features or their components. When constructing Hessian matrix with SURF algorithm, Hessian algorithm is sensitive to gradient change. It can describe the gradient of image well. The method of calculating gradient magnitude before Hessian can effectively improve the effect of existing SURF algorithm. This method solves the problem of small number of feature points and non-uniform feature points in traditional SURF algorithm. It has the advantages of high accuracy of feature points extraction and better noise suppression. It can be introduced into underwater three-dimensional reconstruction, which can effectively improve the accuracy and quality of underwater target three-dimensional reconstruction and provide strong support for underwater observation and operation of underwater vehicles.

Underwater image enhancement technology is one of the key parts of underwater image processing, Since under water environment has greater attenuation and scattering effects on reflected light than the onshore environment that lead to uneven imaging, low contrast, color decay, etc. The underwater enhancement process is intended to highlight some key information of the image and eliminate the irrelevant information so as to make the preparations for subsequent scene segmentation, target recognition, target tracking and predictive analysis and processing. In this paper, we will focus on the problem of underwater image with low quality, and analyze the typical methods of box-based image enhancement methods. At the same time, Retinex image enhancement method based on optical analysis was introduced, and this paper proposed an improved enhancement method based on the Retinex methods. First we convert the image into HSV space for saliency analysis, next, learn the parameters of the harmonic HSV space by statistical learning. And then, we use the weights obtained by learning to predict the enhancement parameter for the decay image. Finally, the experiment verified that the method had better adaptability and better effect over the traditional Retinex algorithm in underwater image enhancement.

The application of deep learning sonar target detection is severely limited due to the small amount of sonar images, especially for submarine shipwreck. Aiming to overcome the over-fit of training problem and improve accuracy of detection, we proposed a method which combine deep generation networks and transfer learning for sonar shipwrecks detection. Specifically, in deep generation network, we used similarity measurement to improved optimization, which generate high quality fake image and laid the further foundation of data. Then, in transfer learning detection, we used multi-layer adaptation and multi-core MMD to fine-tune and frozen pre-trained model, prevent the problem of over-fit and improve the generalization and stability of the system. And we combined the methods of regional suggestion and regression for target detection to guarantee precision of detection. Finally, the contrast experiment of sonar shipwrecks is carried out the effectiveness of the proposed method.

Structural and functional alterations in the inferior parietal lobule (IPL) in schizophrenia have been frequently reported; however, the IPL connectivity changes in schizophrenia remain largely unknown. Based on heterogeneity of the IPL in structure, connection and function, we hypothesize that the resting-state functional connectivities (rsFCs) of the IPL subregions are differentially affected in schizophrenia. This study included 95 schizophrenia patients and 104 healthy controls. The IPL subregions were defined according to a previous in vivo connection-based parcellation study. We calculated the rsFC of each IPL subregion and compared them between the two groups while controlling for the effects of age, gender, and grey matter volume. Among the six subregions of the left IPL and the five subregions of the right IPL, only the bilateral PFm (a transition zone of the IPL) subregions exhibited abnormal rsFC in schizophrenia. Specifically, the left PFm showed increased rsFC with the bilateral lingual gyri in schizophrenia patients than in healthy controls. The right PFm exhibited increased rsFC with the right lingual gyrus and inferior occipital gyrus, and bilateral mid-cingulate and sensorimotor cortices in schizophrenia patients. These findings suggest a selective rsFC abnormality in the IPL subregions in schizophrenia, characterized by the increased rsFC between the PFm subregion of the IPL and the visual and sensorimotor areas., Highlights • Investigation of the subregional-level rsFC changes of the IPL in schizophrenia • The PFm is only IPL subregion with significant rsFC changes in schizophrenia. • Suggesting a selective rsFC abnormality in the IPL subregions in schizophrenia

in the field of ocean development, fish swarm detection has significance for AUV’s autonomous navigation and fishing industry. Aim to present fish swarm detections are common based on $2D$ which lack of spatial information, has low accuracy and bad real-time performance, so we proposed systematic fish swarm detection and position method. We used deep learning target detection system to detect fish and used binocular vision position system to position, then fused every fish’s $3D$ information in camera vision to displayed fish swarm spatial information through radar map. Finally, the contrast experiment and is carried out to verify the effectiveness of the proposed method.

The sortilin-related receptor 1 (SORL1) gene has been associated with late-onset Alzheimer's disease (LOAD) and structural impairments in several ethnic populations. However, how this gene affects brain function properties remains unclear. We investigated associations of SORL1 rs2070045 with functional connectivity density (FCD) in healthy young adults. This single-nucleotide polymorphism was selected because it is the SORL1 variant that has been frequently associated with LOAD in several populations, including the Chinese Han population. A total of 275 healthy young Chinese Han subjects with successful genotyping and MRI examinations were included. The effect of SORL1 rs2070045 was explored using a voxel-wise FCD analysis. A significant effect of SORL1 rs2070045 on the FCD was found in the right inferior temporal gyrus. The risk G allele carriers of the rs2070045 exhibited a lower FCD than the protective TT carriers. This effect was independent of the status of apolipoprotein E. This study provides the first evidence that the SORL1 gene is associated with brain FCD differences in healthy young adults. In the genetic risk subjects, connectivity impairment already starts during young adulthood, which may predispose the risk allele carriers to be susceptible to LOAD after several decades.

Small sample sizes and large inter-subject variations result in inconsistent findings in resting-state cerebral blood flow (CBF) in schizophrenia. The CBF connectivity alterations in schizophrenia remain unclear. Recently, three-dimensional pseudo-continuous arterial spin labeling (pcASL) imaging was performed to measure the resting-state CBF in 100 schizophrenia patients and 94 healthy comparison subjects. The normalized CBF was used to reduce the inter-subject variations. Both group comparisons in the CBF and correlations between the CBF alterations and clinical parameters were assessed. The CBF connectivity of the brain regions with regional CBF differences was also compared between the groups. Compared with the healthy controls, the schizophrenia patients exhibited increased CBF in the bilateral inferior temporal gyri, thalami and putamen and decreased CBF in the left insula and middle frontal gyrus and the bilateral anterior cingulate cortices and middle occipital gyri. In the schizophrenia patients, significant correlations were identified between the CBF and clinical parameters. Importantly, the schizophrenia patients exhibited CBF disconnections between the left thalamus and right medial superior frontal gyrus and between the left insula and left postcentral gyrus. Our results suggest that schizophrenia patients may exhibit both regional CBF abnormalities and deficits in CBF connectivity, which may underlie the clinical symptoms of schizophrenia.

Diverse brain structural and functional changes have been reported in schizophrenia. Identifying different types of brain changes may help to understand the neural mechanisms and to develop reliable biomarkers in schizophrenia. We aimed to categorize different grey matter changes in schizophrenia based on grey matter volume (GMV) and cerebral blood flow (CBF). Structural and perfusion magnetic resonance imaging data were acquired in 100 schizophrenia patients and 95 healthy comparison subjects. Voxel-based GMV comparison was used to show structural changes, CBF analysis was used to demonstrate functional changes. We identified three types of grey matter changes in schizophrenia: structural and functional impairments in the anterior cingulate cortex and insular cortex, displaying reduction in both GMV and CBF; structural impairment with preserved function in the frontal and temporal cortices, demonstrating decreased GMV with normal CBF; pure functional abnormality in the anterior cingulate cortex and lateral prefrontal cortex and putamen, showing altered CBF with normal GMV. By combination of GMV and CBF, we identified three types of grey matter changes in schizophrenia. These findings may help to understand the complex manifestations and to develop reliable biomarkers in schizophrenia.

Delusions are cardinal positive symptoms in schizophrenia; however, the neural substrates of delusions remain unknown. In the present study, we investigated the neural correlates of delusions in schizophrenia using multi-modal magnetic resonance imaging (MRI) techniques. Diffusion, structural and perfusion MRIs were performed in 19 schizophrenia patients with severe delusions, 30 patients without delusions and 30 healthy controls. Fractional anisotropy (FA), gray matter volume (GMV) and cerebral blood flow (CBF) were voxel-wisely compared among the three groups. Although patients without delusions exhibited decreased FA in white matter regions and decreased GMV in gray matter regions relative to controls, patients with severe delusions demonstrated comparable FA in all of these white matter regions and similar GMV in most of these gray matter regions. Both patient subgroups had less GMV in the amygdala and anterior cingulate cortex than controls. Although two patient subgroups showed consistent CBF changes relative to controls, only CBF in the anterior cingulate cortex was lower in patients with severe delusions than in patients without delusions. These findings suggest that schizophrenia patients with severe delusions have relatively normal structural integrity. Importantly, the excessively reduced perfusion in the anterior cingulate cortex may be associated with the development of delusions in schizophrenia.

Clinical and brain structural differences have been reported between patients with familial and sporadic schizophrenia; however, little is known about the brain functional differences between the two subtypes of schizophrenia. Twenty-six patients with familial schizophrenia (PFS), 26 patients with sporadic schizophrenia (PSS) and 26 healthy controls (HC) underwent a resting-state functional magnetic resonance imaging. The whole-brain functional network was constructed and analyzed using graph theoretical approaches. Topological properties (including global, nodal and edge measures) were compared among the three groups. We found that PFS, PSS and HC exhibited common small-world architecture of the functional brain networks. However, at a global level, only PFS showed significantly lower normalized clustering coefficient, small-worldness and local efficiency, indicating a randomization shift of their brain networks. At a regional level, PFS and PSS disrupted different neural circuits, consisting of abnormal nodes (increased or decreased nodal centrality) and edges (decreased functional connectivity strength), which were widely distributed throughout the entire brain. Furthermore, some of these altered network measures were significantly correlated with severity of psychotic symptoms. These results suggest that familial and sporadic schizophrenia had segregated disruptions in the topological organization of the intrinsic functional brain network, which may be due to different etiological contributions.

Although extensive resting-state functional connectivity (rsFC) changes have been reported in schizophrenia, rsFC changes in the temporal pole (TP) remain unknown. The TP contains several subregions with different connection patterns; however, it is not known whether TP subregions are differentially affected in schizophrenia. Sixty-six schizophrenia patients and 76 healthy comparison subjects underwent resting-state fMRI using a sensitivity-encoded spiral-in (SENSE-SPIRAL) imaging sequence to reduce susceptibility-induced signal loss and distortion. The TP was subdivided into the dorsal (TPd) and ventral (TPv) subregions. Mean fMRI time series were extracted for each TP subregion and entered into a seed-based rsFC analysis. Direct between-group comparisons revealed reduced rsFC between the right TPd and brain regions involved in language processing and multisensory integration in schizophrenia, including the left superior temporal gyrus, left mid-cingulate cortex and right insular cortex. The rsFC changes of the right TPd in schizophrenia were independent of the grey matter reduction of this subregion. Moreover, these rsFC changes were unrelated to illness severity, duration of illness and antipsychotic medication dosage. No significant group differences were observed in the rsFC of the left TPd and bilateral TPv subregions. These findings suggest a selective (the right TPd) functional disconnection of TP subregions in schizophrenia.

Schizophrenia patients have shown altered resting-state functional connectivity (rsFC) of the cingulate cortex; however, it is unknown whether rsFCs of the cingulate subregions are differentially affected in this disorder. We aimed to clarify the issue by comparing rsFCs of each cingulate subregion between healthy controls and schizophrenia patients. A total of 102 healthy controls and 94 schizophrenia patients underwent resting-state functional magnetic resonance imaging with a sensitivity-encoded spiral-in imaging sequence to reduce susceptibility-induced signal loss and distortion. The cingulate cortex was divided into nine subregions, including the subgenual anterior cingulate cortex (ACC), areas 24 and 32 of the pregenual ACC, areas 24 and 32 of the anterior mid-cingulate cortex (aMCC), posterior MCC (pMCC), dorsal (dPCC) and ventral (vPCC) posterior cingulate cortex (PCC) and retrosplenial cortex (RSC). The rsFCs of each cingulate subregion were compared between the two groups and the atrophy effect was considered. Results with and without global signal regression were reported. Most cingulate subregions exhibited decreased rsFCs in schizophrenia after global signal regression (GSR). Without GSR, only increased rsFC was found in schizophrenia, which primarily restricted to the aMCC, PCC and RSC. Some of these increased rsFCs were also significant after GSR. These findings suggest that GSR can greatly affect between-group differences in rsFCs and the consistently increased rsFCs may challenge the functional disconnection hypothesis of schizophrenia.

Apolipoprotein E (APOE) and sortilin-related receptor (SORL1) genes act on the same metabolic pathway and have been associated with Alzheimer’s disease (AD) characterized by hippocampal impairment. Although the effects of APOE on hippocampal resting-state functional connectivity (rsFC) have been reported, the main effects of SORL1 and SORL1 × APOE interactions on hippocampal rsFC in healthy subjects remain largely unknown. Here, we systematically investigated the main effects of SORL1 rs2070045, and APOE, and their interaction effects on hippocampal rsFC in healthy young adults. The main effect of APOE showed that risk ε4 carriers had decreased positive hippocampal rsFC with the precuneus/posterior cingulate cortex and subgenual anterior cingulate cortex, and increased positive hippocampal rsFC with the sensorimotor cortex compared with non-ε4 carriers. The main effect of SORL1 showed that risk G-allele carriers had decreased positive rsFC between the hippocampus and middle temporal gyrus compared with TT carriers. No significant additive interaction was observed. Instead, significant SORL1 × APOE non-additive interaction was found in negative rsFC between the hippocampus and inferior frontal gyrus. Compared with subjects with TT genotype, SORL1 G-allele carriers had a stronger negative rsFC in APOE ε4 carriers, but a weaker negative rsFC in APOE non-ε4 carriers. These findings suggest that SORL1 and APOE genes modulate different hippocampal rsFCs and have a complex interaction. The SORL1- and APOE-dependent hippocampal connectivity changes may at least partly account for their association with AD. Electronic supplementary material The online version of this article (doi:10.1007/s00429-017-1377-3) contains supplementary material, which is available to authorized users.