While convolutional neural network (CNN) has been demonstrating powerful ability to learn hierarchical spatial features from medical images, it is still difficult to apply it directly to restingstate functional MRI (rs-fMRI) and the derived brain functional networks (BFNs). We propose a novel CNN framework to simultaneously learn embedded features from BFNs for brain disease diagnosis. Since BFNs can be built by considering both static and dynamic functional connectivity (FC), we first decompose rs-fMRI into multiple static BFNs with modified independent component analysis. Then, voxel-wise variability in dynamic FC is used to quantify BFN dynamics. A set of paired 3D images representing static/dynamic BFNs can be fed into 3D CNNs, from which we can hierarchically and simultaneously learn static/dynamic BFN features. As a result, dynamic BFN features can complement static BFN features and, at meantime, different BFNs can help each other towards a joint and better classification. We validate our method with a publicly accessible, large cohort of rs-fMRI dataset in earlystage mild cognitive impairment (eMCI) diagnosis, which is one of the most challenging problems to the clinicians. By comparing with a conventional method, our method shows significant diagnostic performance improvement by almost 10%. This result demonstrates the effectiveness of deep learning in preclinical Alzheimer's disease diagnosis, based on the complex and highdimensional voxel-wise spatiotemporal patterns of the resting-state brain functional connectomics. The framework provides a new but intuitive way to fully exploit deeply embedded diagnostic features from rs-fMRI for better individualized diagnosis of various neurological diseases.

Radiation therapy, a major method of treatment for brain cancer, may cause severe brain injuries after many years. We used a rare and unique cohort of nasopharyngeal carcinoma patients with normal-appearing brains to study possible early irradiation injury in its presymptomatic phase before severe, irreversible necrosis happens. The aim is to detect any structural or functional imaging biomarker that is sensitive to early irradiation injury, and to understand the recovery and progression of irradiation injury that can shed light on outcome prediction for early clinical intervention. We found an acute increase in local brain activity that is followed by extensive reductions in such activity in the temporal lobe and significant loss of functional connectivity in a distributed, largescale, high-level cognitive function-related brain network. Intriguingly, these radiosensitive functional alterations were found to be fully or partially recoverable. In contrast, progressive late disruptions to the integrity of the related far-end white matter structure began to be significant after one year. Importantly, early increased local brain functional activity was predictive of severe later

Major depressive disorder (MDD) is a serious mental illness characterized by dysfunctional connectivity among distributed brain regions. Previous connectome studies based on functional magnetic resonance imaging (fMRI) have focused primarily on undirected functional connectivity and existing directed effective connectivity (EC) studies concerned mostly task-based fMRI and incorporated only a few brain regions. To overcome these limitations and understand whether MDD is mediated by within-network or between-network connectivities, we applied spectral dynamic causal modeling to estimate EC of a large-scale network with 27 regions of interests from four distributed functional brain networks (default mode, executive control, salience, and limbic networks), based on large sample-size resting-state fMRI consisting of 100 healthy subjects and 100 individuals with first-episode drug-naive MDD. We applied a newly developed parametric empirical Bayes (PEB) framework to test specific hypotheses. We showed that MDD altered EC both within and between high-order functional networks. Specifically, MDD is associated with reduced excitatory connectivity mainly within the default mode network (DMN), and between the default mode and salience networks. In addition, the network-averaged inhibitory EC within the DMN was found to be significantly elevated in the MDD. The coexistence of the reduced excitatory but increased inhibitory causal connections within the DMNs may underlie disrupted self-recognition and emotional control in MDD.Guoshi Li and Yujie Liu contributed equally to this study. Hum Brain Mapp. 2020;41:865-881. wileyonlinelibrary.com/journal/hbm 865 SUPPORTING INFORMATION Additional supporting information may be found online in the Supporting Information section at the end of this article. How to cite this article: Li G, Liu Y, Zheng Y, et al. Largescale dynamic causal modeling of major depressive disorder based on resting-state functional magnetic resonance imaging.

High-grade glioma (HGG) is a lethal cancer with poor outcome. Accurate preoperative overall survival (OS) time prediction for HGG patients is crucial for treatment planning. Traditional presurgical and noninvasive OS prediction studies have used radiomics features at the local lesion area based on the magnetic resonance images (MRI). However, the highly complex lesion MRI appearance may have large individual variability, which could impede accurate individualized OS prediction. In this paper, we propose a novel concept, namely brain connectomics-based OS prediction. It is based on presurgical resting-state functional MRI (rs-fMRI) and the non-local, large-scale brain functional networks where the global and systemic prognostic features rather than the local lesion appearance are used to predict OS. We propose that the connectomics features could capture tumor-induced network-level alterations that are associated with prognosis. We construct both low-order (by means of sparse representation with regional rs-fMRI signals) and high-order functional connectivity (FC) networks (characterizing more complex multi-regional relationship by synchronized dynamics FC time courses). Then, we conduct a graph-theoretic analysis on both networks for a jointly, machine-learning-based individualized OS prediction. Based on a preliminary dataset (N = 34 with bad OS, mean OS, ~400 days; N = 34 with good OS, mean OS, ~1030 days), we achieve a promising OS prediction accuracy (86.8%) on separating the individuals with bad OS from those with good OS. However, if using only conventionally derived descriptive features (e.g., age and tumor characteristics), the accuracy is low (63.2%). Our study highlights the importance of the rs-fMRI and brain functional connectomics for treatment planning.

No abstract