Pulvinar Response Profiles and Connectivity Patterns to Object Domains.

Distributed cortical regions show differential responses to visual objects belonging to different domains varying by animacy (e.g., animals vs tools), yet it remains unclear whether this is an organization principle also applying to the subcortical structures. Combining multiple fMRI activation experiments (two main experiments and six validation datasets; 12 females and 9 males in the main Experiment 1; 10 females and 10 males in the main Experiment 2), resting-state functional connectivity, and task-based dynamic causal modeling analysis in human subjects, we found that visual processing of images of animals and tools elicited different patterns of response in the pulvinar, with robust left lateralization for tools, and distinct, bilateral (with rightward tendency) clusters for animals. Such domain-preferring activity distribution in the pulvinar was associated with the magnitude with which the voxels were intrinsically connected with the corresponding domain-preferring regions in the cortex. The pulvinar-to-right-amygdala path showed a one-way shortcut supporting the perception of animals, and the modulation connection from pulvinar to parietal showed an advantage to the perception of tools. These results incorporate the subcortical regions into the object processing network and highlight that domain organization appears to be an overarching principle across various processing stages in the brain.SIGNIFICANCE STATEMENT Viewing objects belonging to different domains elicited different cortical regions, but whether the domain organization applied to the subcortical structures (e.g., pulvinar) was unknown. Multiple fMRI activation experiments revealed that object pictures belonging to different domains elicited differential patterns of response in the pulvinar, with robust left lateralization for tool pictures, and distinct, bilateral (with rightward tendency) clusters for animals. Combining the resting-state functional connectivity and dynamic causal modeling analysis on task-based fMRI data, we found domain-preferring activity distribution in the pulvinar aligned with that in cortical regions. These results highlight the need for coherent visual theories that explain the mechanisms underlying the domain organization across various processing stages.

Pain-Discriminating Information Decoded From Spatiotemporal Patterns of Hemodynamic Responses Measured by fMRI in the Human Brain.

Functional magnetic resonance imaging (fMRI) has been widely used in studying the neural mechanisms of pain in the human brain, primarily focusing on where in the brain pain-elicited neural activities occur (i.e., the spatial distribution of pain-related brain activities). However, the temporal dynamics of pain-elicited hemodynamic responses (HDRs) measured by fMRI may also contain information specific to pain processing but have been largely neglected. Using high temporal resolution fMRI (TR = 0.8 s) data acquired from 62 healthy participants, in the present study we aimed to test whether pain-distinguishing information could be decoded from the spatial pattern of the temporal dynamics (i.e., the spatiotemporal pattern) of HDRs elicited by painful stimuli. Specifically, the peak latency and the response duration were used to characterize the temporal dynamics of HDRs to painful laser stimuli and non-painful electric stimuli, and then were compared between the two conditions (i.e., pain and no-pain) using a voxel-wise univariate analysis and a multivariate pattern analysis. Furthermore, we also tested whether the two temporal characteristics of pain-elicited HDRs and their spatial patterns were associated with pain-related behaviors. We found that the spatial patterns of HDR peak latency and response duration could successfully discriminate pain from no-pain. Interestingly, we also observed that the Pain Vigilance and Awareness Questionnaire (PVAQ) scores were correlated with the average response duration in bilateral insula and secondary somatosensory cortex (S2) and could also be predicted from the across-voxel spatial patterns of response durations in the middle cingulate cortex and middle frontal gyrus only during painful condition but not during non-painful condition. These findings indicate that the spatiotemporal pattern of pain-elicited HDRs may contain pain-specific information and highlight the importance of studying the neural mechanisms of pain by taking advantage of the high sensitivity of fMRI to both spatial and temporal information of brain responses.

[Characterization of brain deactivations elicited by transient painful and tactile stimuli using functional MRI].

The aim of the present study was to explore the specific pattern of brain deactivation elicited by painful stimuli, in contrast with that elicited by tactile stimuli. Functional magnetic resonance imaging (fMRI) data were collected from 62 healthy subjects under painful and tactile stimuli with varying intensities. The brain deactivations under different conditions were identified using the general linear model. Two-way analysis of variance (ANOVA) was performed to test whether there was a significant interaction between perceived stimulus intensity (factor 1: high intensity, low intensity) and stimulus modality (factor 2: pain, touch) on the brain deactivations. The results showed that there were significant interactions between stimulus intensity and stimulus modality on the deactivations of left medial superior frontal gyrus, left middle occipital gyrus, left superior frontal gyrus and right middle occipital gyrus (P < 0.05, Cluster-level FWE). The deactivations induced by painful stimuli with low perceived intensity (ß = -3.38 ± 0.52) were significantly stronger than those induced by painful stimuli with high perceived intensity (ß = -1.22 ± 0.54) (P < 0.001), whereas the differences between the deactivations induced by tactile stimuli with different perceived intensities were not statistically significant. In addition, there were no significant differences between the deactivations elicited by painful and tactile stimuli with the same stimulus intensities. These results suggest that there is a specific relationship between the deactivations induced by painful stimuli in multiple brain regions (such as the left medial superior frontal gyrus) and the stimulus intensity, providing evidence for a deeper understanding of the brain mechanisms underlying pain perception.

Brain-activation-based individual identification reveals individually unique activation patterns elicited by pain and touch.

Pain is subjective and perceived differently in different people. However, individual differences in pain-elicited brain activations are largely overlooked and often discarded as noises. Here, we used a brain-activation-based individual identification procedure to investigate the uniqueness of the activation patterns within the whole brain or brain regions elicited by nociceptive (laser) and tactile (electrical) stimuli in each of 62 healthy participants. Specifically, brain activation patterns were used as "fingerprints" to identify each individual participant within and across sensory modalities, and individual identification accuracy was calculated to measure each individual's identifiability. We found that individual participants could be successfully identified using their brain activation patterns elicited by nociceptive stimuli, tactile stimuli, or even across modalities. However, different participants had different identifiability; importantly, the within-pain, but not within-touch or cross-modality, individual identifiability obtained from three brain regions (i.e., the left superior frontal gyrus, the middle temporal gyrus and the insular gyrus) were inversely correlated with the scores of Pain Vigilance and Awareness Questionnaire (i.e., how a person is alerted to pain) across participants. These results suggest that each individual has a unique pattern of brain responses to nociceptive stimuli which contains both modality-nonspecific and pain-specific information and may be associated with pain-related behaviors shaped by his/her own personal experiences and highlight the importance of a transition from group-level to individual-level characterization of brain activity in neuroimaging studies.

Functional MRI evidence for primary motor cortex plasticity contributes to the disease's severity and prognosis of cervical spondylotic myelopathy patients.

OBJECTIVE: To investigate the brain mechanism of non-correspondence between diseases severity and compression degree of the spinal cord in cervical spondylotic myelopathy (CSM) patients and to test the utility of brain imaging biomarkers for predicting prognosis of CSM. METHODS: We calculated voxel-wise zALFF from 54 CSM patients and 50 healthy controls using resting-state fMRI data. In analysis 1, we identified the brain regions exhibited significant differences of zALFF between CSM patients and healthy controls. In analyses 2 through 3, we investigated the zALFF differences between light-symptom CSM patients and severe-symptom CSM patients while carefully matching the degree of compression between these two groups. In analysis 4, we tested the utility of zALFF within the primary motor cortex (M1) for predicting the prognosis of CSM. RESULTS: We found that (1) compared with the healthy controls, CSM patients exhibited higher ALFF within left M1, bilateral superior frontal gyrus, and lower zALFF within right precuneus and calcarine, suggesting altered brain neural activity in CSM patients; (2) after matching the compression degree, the CSM patients with more severe clinical symptoms exhibited higher zALFF within M1, indicating cortical function contributes to disease's severity of CSM; (3) taking the M1 zALFF as features in the prognosis prediction model improves the prediction accuracy, indicating that the M1 zALFF provide additional value for predicting the prognosis of CSM patients following decompression surgery. CONCLUSION: The functional state of M1 contributes to the disease's severity of CSM and can provide complementary information for predicting the prognosis of CSM following decompression surgery. KEY POINTS: • Cervical spondylotic myelopathy (CSM) patients exhibited increased zALFF within the primary motor cortex (M1), bilateral superior frontal gyrus, and decreased zALFF within the right precuneus and calcarine. • After matching the compression degree, the CSM patients with more severe clinical symptoms exhibited higher zALFF within M1, indicating cortical function contributes to disease severity of CSM. • zALFF within M1 provided additional value for predicting the prognosis of CSM patients.

Feedforward and feedback pathways of nociceptive and tactile processing in human somatosensory system: A study of dynamic causal modeling of fMRI data.

Nociceptive and tactile information is processed in the somatosensory system via reciprocal (i.e., feedforward and feedback) projections between the thalamus, the primary (S1) and secondary (S2) somatosensory cortices. The exact hierarchy of nociceptive and tactile information processing within this 'thalamus-S1-S2' network and whether the processing hierarchy differs between the two somatosensory submodalities remains unclear. In particular, two questions related to the ascending and descending pathways have not been addressed. For the ascending pathways, whether tactile or nociceptive information is processed in parallel (i.e., 'thalamus-S1' and 'thalamus-S2') or in serial (i.e., 'thalamus-S1-S2') remains controversial. For the descending pathways, how corticothalamic feedback regulates nociceptive and tactile processing also remains elusive. Here, we aimed to investigate the hierarchical organization for the processing of nociceptive and tactile information in the 'thalamus-S1-S2' network using dynamic causal modeling (DCM) combined with high-temporal-resolution fMRI. We found that, for both nociceptive and tactile information processing, both S1 and S2 received inputs from thalamus, indicating a parallel structure of ascending pathways for nociceptive and tactile information processing. Furthermore, we observed distinct corticothalamic feedback regulations from S1 and S2, showing that S1 generally exerts inhibitory feedback regulation independent of external stimulation whereas S2 provides additional inhibition to the thalamic activity during nociceptive and tactile information processing in humans. These findings revealed that nociceptive and tactile information processing have similar hierarchical organization within the somatosensory system in the human brain.

Stability test of canonical correlation analysis for studying brain-behavior relationships: The effects of subject-to-variable ratios and correlation strengths.

Canonical correlation analysis (CCA), a multivariate approach to identifying correlations between two sets of variables, is becoming increasingly popular in neuroimaging studies on brain-behavior relationships. However, the CCA stability in neuroimaging applications has not been systematically investigated. Although it is known that the number of subjects should be greater than the number of variables due to the curse of dimensionality, it is unclear at what subject-to-variable ratios (SVR) and at what correlation strengths the CCA stability can be maintained. Here, we systematically assessed the CCA stability, in the context of investigating the relationship between the brain structural/functional imaging measures and the behavioral measures, by measuring the similarity of the first-mode canonical variables across randomly sampled subgroups of subjects from a large set of 936 healthy subjects. Specifically, we tested how the CCA stability changes with SVR under two different brain-behavior correlation strengths. The same tests were repeated using an independent data set (n = 700) for validation. The results confirmed that both SVR and correlation strength affect greatly the CCA stability-the CCA stability cannot be guaranteed if the SVR is not sufficiently high or the brain-behavior relationship is not sufficiently strong. Based on our quantitative characterization of CCA stability, we provided a practical guideline to help correct interpretation of CCA results and proper applications of CCA in neuroimaging studies on brain-behavior relationships.

Prediction of mediastinal lymph node metastasis based on 18F-FDG PET/CT imaging using support vector machine in non-small cell lung cancer.

OBJECTIVE: The purpose of this study was to develop a classification method based on support vector machine (SVM) to improve the diagnostic performance of 18F-fluorodeoxyglucose (FDG) positron emission tomography/computed tomography (PET/CT) to detect the lymph node (LN) metastasis in non-small cell lung cancer (NSCLC). METHOD: Two hundred nineteen lymph nodes (37 metastatic) from 71 patients were evaluated in this study. SVM models were developed with 7 LN features. The area under the curve (AUC) and accuracy of 9 models were compared to select the best model. The best SVM model was simplified on the basis of the feature weights and value distribution to further suit the clinical application. RESULTS: The maximum, minimum, and mean accuracy of the best model was 91.89% (68/74, 95% CI 83.11~96.54%), 66.22% (49/74, 95% CI 54.85~75.98%), and 80.09% (59,266/74,000, 95% CI 70.27~89.19%), respectively, with an AUC of 0.94, 0.66, and 0.81, respectively. The best SVM model was finally simplified into a score rule: LNs with scores more than 3.0 were considered as malignant ones, whereas LNs with scores less than 1.5 tended to be benign ones. For the LNs with scores within a range of 1.5-3.0, metastasis was suspected. CONCLUSION: An SVM model based on 18F-FDG PET/CT images was able to predict the metastatic LNs for patients with NSCLC. The ratio of the maximum of standard uptake value of LNs to aortic arch played a major role in the model. After simplification, the model could be transferred into a scoring method which may partly help clinicians determine the clinical staging of patients with NSCLC relatively easier. KEY POINTS: • The SVM model based on 18F-FDG PET/CT features may help clinicians to make a decision for metastatic mediastinal lymph nodes in patients with NSCLC. • The SURblood plays a major role in the SVM model. • The score rule based on the SVM model simplified the complexity of the model and may partly help clinicians determine the clinical staging of patients with NSCLC relatively easier.

Influenza H5/H7 Virus Vaccination in Poultry and Reduction of Zoonotic Infections, Guangdong Province, China, 2017-18.

We compared the detection frequency of avian influenza H7 subtypes at live poultry markets in Guangdong Province, China, before and after the introduction of a bivalent H5/H7 vaccine in poultry. The vaccine was associated with a 92% reduction in H7 positivity rates among poultry and a 98% reduction in human H7N9 cases.

Molecular Evolution, Diversity, and Adaptation of Influenza A(H7N9) Viruses in China.

The substantial increase in prevalence and emergence of antigenically divergent or highly pathogenic influenza A(H7N9) viruses during 2016-17 raises concerns about the epizootic potential of these viruses. We investigated the evolution and adaptation of H7N9 viruses by analyzing available data and newly generated virus sequences isolated in Guangdong Province, China, during 2015-2017. Phylogenetic analyses showed that circulating H7N9 viruses belong to distinct lineages with differing spatial distributions. Hemagglutination inhibition assays performed on serum samples from patients infected with these viruses identified 3 antigenic clusters for 16 strains of different virus lineages. We used ancestral sequence reconstruction to identify parallel amino acid changes on multiple separate lineages. We inferred that mutations in hemagglutinin occur primarily at sites involved in receptor recognition or antigenicity. Our results indicate that highly pathogenic strains likely emerged from viruses circulating in eastern Guangdong Province during March 2016 and are associated with a high rate of adaptive molecular evolution.

Sky Detection in Hazy Image.

Sky detection plays an essential role in various computer vision applications. Most existing sky detection approaches, being trained on ideal dataset, may lose efficacy when facing unfavorable conditions like the effects of weather and lighting conditions. In this paper, a novel algorithm for sky detection in hazy images is proposed from the perspective of probing the density of haze. We address the problem by an image segmentation and a region-level classification. To characterize the sky of hazy scenes, we unprecedentedly introduce several haze-relevant features that reflect the perceptual hazy density and the scene depth. Based on these features, the sky is separated by two imbalance SVM classifiers and a similarity measurement. Moreover, a sky dataset (named HazySky) with 500 annotated hazy images is built for model training and performance evaluation. To evaluate the performance of our method, we conducted extensive experiments both on our HazySky dataset and the SkyFinder dataset. The results demonstrate that our method performs better on the detection accuracy than previous methods, not only under hazy scenes, but also under other weather conditions.

Assessing the risk of downwind spread of avian influenza virus via airborne particles from an urban wholesale poultry market.

Interspecies transmissions of avian influenza viruses (AIV) occur at the human-poultry interface, among which the live poultry markets (LPMs) are easily assessed by urban residents. Thousands of live poultry from different farms arrive daily at wholesale markets before being sold to retail markets. We assessed the risk of AIV downwind spread via airborne particles from a representative wholesale market in Guangzhou. Air samples were collected using the cyclone-based NIOSH bioaerosol samplers at different locations inside a wholesale market, and viral RNA and avian 18S RNA were quantified using quantitative real-time RT-PCR. Computational Fluid Dynamics (CFD) modeling was performed to investigate the AIV spread pattern. Viral RNA was readily detected from 19 out of 21 air sampling events, predominantly from particles larger than 1 µm. The concentration of viral RNA detected at the poultry holding area was 4.4 × 105 copies/m3 and was as high as 2.6 × 104 copies/m3 100 m downwind. A high concentration of avian 18S RNA (2.5 × 108 copies/m3) detected at the poultry holding area was used for assessing the potential spread of avian influenza virus during outbreak situations. CFD modeling indicated the combined effect of wind direction and surrounding buildings on the spread of virus and a slow decay rate of the virus in the air in the downwind direction. Because of the large volume of poultry trade daily, wholesale markets located in urban areas may pose considerable AIV infection risk to neighboring residents via wind spread, even in the absence of direct contact with poultry.

A cluster of coxsackievirus A21 associated acute respiratory illness: the evidence of efficient transmission of CVA21.

In March 2016, a cluster of unexplained respiratory illnesses was reported by the acute respiratory infections (ARI) surveillance system of Guangdong Province, China. Twenty-three high school students and one teacher from the four neighboring classes were admitted to a hospital. CVA21 was found in eight of fourteen patients. Phylogenetic analysis suggested that the CVA21 outbreak was most likely caused by transmission of the virus from person to person. This is the first report of an ARI outbreak caused by CVA21, which suggests that CVA21 has the potential to be transmitted efficiently from person to person and should be closely monitored by clinicians and public health agencies.

Effect of Live Poultry Market Interventions on Influenza A(H7N9) Virus, Guangdong, China.

Since March 2013, three waves of human infection with avian influenza A(H7N9) virus have been detected in China. To investigate virus transmission within and across epidemic waves, we used surveillance data and whole-genome analysis of viruses sampled in Guangdong during 2013-2015. We observed a geographic shift of human A(H7N9) infections from the second to the third waves. Live poultry market interventions were undertaken in epicenter cities; however, spatial phylogenetic analysis indicated that the third-wave outbreaks in central Guangdong most likely resulted from local virus persistence rather than introduction from elsewhere. Although the number of clinical cases in humans declined by 35% from the second to the third waves, the genetic diversity of third-wave viruses in Guangdong increased. Our results highlight the epidemic risk to a region reporting comparatively few A(H7N9) cases. Moreover, our results suggest that live-poultry market interventions cannot completely halt A(H7N9) virus persistence and dissemination.

Isolation of H5N6, H7N9 and H9N2 avian influenza A viruses from air sampled at live poultry markets in China, 2014 and 2015.

Zoonotic infections by avian influenza viruses occur at the human-poultry interface, but the modes of transmission have not been fully investigated. We assessed the potential for airborne and fomite transmission at live poultry markets in Guangzhou city and in Hong Kong Special Administrative Region (SAR), China, during 2014 and 2015. Viral genome and infectious avian influenza A viruses of H5N6, H7N9, and H9N2 subtypes were detected predominantly from particles larger or equal to 1 µm in diameter in the air sampled with cyclone-based bioaerosol samplers at the live poultry markets in Guangzhou. Influenza A(H9N2) viruses were ubiquitously isolated every month during the study period from air and environmental swabs, and different lineages of H9N2 virus were isolated from markets where chickens and minor land-based poultry were sold. The use of de-feathering devices increased the quantity of virus-laden airborne particles while market closure reduced the amount of such particles. The results highlight the possibility of airborne transmission of avian influenza viruses among poultry or from poultry to humans within such settings. This may explain epidemiological observations in which some patients with H7N9 infection reported being in markets but no direct contact with live poultry or poultry stalls.

Serologic screenings for H7N9 from three sources among high-risk groups in the early stage of H7N9 circulation in Guangdong Province, China.

BACKGROUND: The aim of this study was to assess the prevalence of the novel avian influenza A virus (H7N9) in three high risk groups. The groups were divided into those exposed through infected individuals, those exposed through poultry and those individuals exposed through the external environment, in the early stage of the epidemic in Guangdong Province, which is located in the southern region of China. METHODS: Serologic studies were conducted among samples collected from individuals who had close contact with the first H7N9 infected patient reported in Guangdong Province, those who were most likely exposed to the first group of H7N9 infected poultry, and those who might have been exposed to H7N9 in the environmental settings, namely hemagglutinin inhibition (HI) and microneutralizaiton(MN) assays using three viruses as antigens. RESULTS: The alignment results of the viral sequences indicated the similarity of the HA gene sequence among viruses from exposure to infected poultry, infected humans and contaminated environments were highly conserved. Seven samples of individuals exposed to contaminated environments were positive in the HI assay and one sample among them was positive in the MN assay using poultry H7N9 virus as the antigen. One sample was positive against human H7N9 virus and 3 samples were positive against environmental H7N9 among those that were in contact with infected patients in HI assay. None of these were positive in MN assay. All HI titers of the 240 samples from those individuals in contact with infected poultry were less than 40 aganist the antigens from three viruses. CONCLUSIONS: The results suggest that when the H7N9 virus was in the early stages of circulation in Guangdong Province, the antigenic sites of the HA proteins of the H7N9 strain isolated from different hosts were highly conserved. The risk of new infection is low in individuals who have contact with the infected patients, poultry or a contaminated environment in the early stages of the circulation of the H7N9 virus.

Multitrait Genetic Analysis Identifies Novel Pleiotropic Loci for Depression and Schizophrenia in East Asians.

BACKGROUND AND HYPOTHESIS: While genetic correlations, pleiotropic loci, and shared genetic mechanisms of psychiatric disorders have been extensively studied in European populations, the investigation of these factors in East Asian populations has been relatively limited. STUDY DESIGN: To identify novel pleiotropic risk loci for depression and schizophrenia (SCZ) in East Asians. We utilized the most comprehensive dataset available for East Asians and quantified the genetic overlap between depression, SCZ, and their related traits via a multitrait genome-wide association study. Global and local genetic correlations were estimated by LDSC and ρ-HESS. Pleiotropic loci were identified by the multitrait analysis of GWAS (MTAG). STUDY RESULTS: Besides the significant correlation between depression and SCZ, our analysis revealed genetic correlations between depression and obesity-related traits, such as weight, BMI, T2D, and HDL. In SCZ, significant correlations were detected with HDL, heart diseases and use of various medications. Conventional meta-analysis of depression and SCZ identified a novel locus at 1q25.2 in East Asians. Further multitrait analysis of depression, SCZ and related traits identified ten novel pleiotropic loci for depression, and four for SCZ. CONCLUSIONS: Our findings demonstrate shared genetic underpinnings between depression and SCZ in East Asians, as well as their associated traits, providing novel candidate genes for the identification and prioritization of therapeutic targets specific to this population.

Vacancy defect engineered BiVO4 with low-index surfaces for photocatalytic application: a first principles study.

BiVO4 has been widely investigated as a photocatalyst material for water splitting due to its outstanding photocatalytic properties. In order to further improve its photocatalytic efficiency, it is necessary to conduct an in-depth study of improvement strategies, such as defect engineering. By focusing on the (001) and (011) surfaces, we carried out a systematic theoretical research on pristine and defective systems, including Bi, V and O vacancies. Based on density functional theory (DFT), the electronic properties, band alignments and Gibbs free energy of pristine and defective BiVO4 have been analyzed. The electronic structures of the (001) and (011) surfaces show different band gaps, and O vacancies make the BiVO4 become an n-type semiconductor, while Bi and V vacancies tend to form a p-type semiconductor. Moreover, the band edge positions indicate that holes are indeed easily accumulated on the (011) surface while electrons tend to accumulate on (001). However, the (011) surface with Bi and V vacancies does not have enough oxidation potential to oxidize water. The reaction free energy shows that O and Bi vacancies could lower the overpotential to some extent.

Exploring neurometabolic alterations in bipolar disorder with suicidal ideation based on proton magnetic resonance spectroscopy and machine learning technology.

Bipolar disorder (BD) is associated with a high risk of suicide. We used proton magnetic resonance spectroscopy (1H-MRS) to detect biochemical metabolite ratios in the bilateral prefrontal white matter (PWM) and hippocampus in 32 BD patients with suicidal ideation (SI) and 18 BD patients without SI, identified potential brain biochemical differences and used abnormal metabolite ratios to predict the severity of suicide risk based on the support vector machine (SVM) algorithm. Furthermore, we analyzed the correlations between biochemical metabolites and clinical variables in BD patients with SI. There were three main findings: (1) the highest classification accuracy of 88% and an area under the curve of 0.9 were achieved in distinguishing BD patients with and without SI, with N-acetyl aspartate (NAA)/creatine (Cr), myo-inositol (mI)/Cr values in the bilateral PWM, NAA/Cr and choline (Cho)/Cr values in the left hippocampus, and Cho/Cr values in the right hippocampus being the features contributing the most; (2) the above seven features could be used to predict Self-rating Idea of Suicide Scale scores (r = 0.4261, p = 0.0302); and (3) the level of neuronal function in the left hippocampus may be related to the duration of illness, the level of membrane phospholipid catabolism in the left hippocampus may be related to the severity of depression, and the level of inositol metabolism in the left PWM may be related to the age of onset in BD patients with SI. Our results showed that the combination of multiple brain biochemical metabolites could better predict the risk and severity of suicide in patients with BD and that there was a significant correlation between biochemical metabolic values and clinical variables in BD patients with SI.