Search

Major depressive disorder (MDD) is a common mental disorder that typically affects a person's mood, cognition, behavior, and physical health. Resting-state functional magnetic resonance imaging (rs-fMRI) data are widely used for computer-aided diagnosis of MDD. While multi-site fMRI data can provide more data for training reliable diagnostic models, significant cross-site data heterogeneity would result in poor model generalizability. Many domain adaptation methods are designed to reduce the distributional differences between sites to some extent, but usually ignore overfitting problem of the model on the source domain. Intuitively, target data augmentation can alleviate the overfitting problem by forcing the model to learn more generalized features and reduce the dependence on source domain data. In this work, we propose a new augmentation-based unsupervised cross-domain fMRI adaptation (AUFA) framework for automatic diagnosis of MDD. The AUFA consists of 1) a graph representation learning module for extracting rs-fMRI features with spatial attention, 2) a domain adaptation module for feature alignment between source and target data, 3) an augmentation-based self-optimization module for alleviating model overfitting on the source domain, and 4) a classification module. Experimental results on 1,089 subjects suggest that AUFA outperforms several state-of-the-art methods in MDD identification. Our approach not only reduces data heterogeneity between different sites, but also localizes disease-related functional connectivity abnormalities and provides interpretability for the model.

Electrophysiological time series were taken simultaneously in two locations in the hippocampus of a rat brain previously described as receiving innervation from the osmosensitive vasopressinergic neurons of the hypothalamus. A hyperosmotic saline solution injection was administered during the time of the experiment. We analyze the recorded time series using different methods. We detect a modification of the delta and theta oscillations just after the perturbation caused by the injection. We compare the quality and information that each one of the methods exhibit and we analyze the characteristics of the perturbation based on a hypothesis that the strength of the functional connections between the vasopressinergic hypothalamic magnocellular neurons and their target in the hippocampus is modified by the perturbation. We built a model of the hypothetic neural connections and numerically calculate the time series produced by the system when simulating the perturbation caused by the saline injection. The theoretical results resemble the experimental findings concerning the frequency and amplitude alterations of the delta and theta bands., Comment: 11 pages, 12 figures

Quantum coherence, emerging from the 'superposition' of quantum states, is widely used in various information processing tasks. Recently, the resource theory of multilevel quantum coherence is attracting substantial attention. In this paper, we mainly study the deterministic transformations of resource pure states via free operations in the theoretical framework for multilevel coherence. We prove that any two multilevel coherent resource pure states can be interconverted with a nonzero probability via a completely positive and trace non-increasing $k$-coherence-preserving map. Meanwhile, we present the condition of the interconversions of two multilevel coherent resource pure states under $k$-coherence-preserving operations. In addition, we obtain that in the resource-theoretic framework of multilevel coherence, no resource state is isolated, that is, given a multilevel coherent pure state $|\psi\rangle$, there exists another multilevel coherent pure state $|\phi\rangle$ and a $k$-coherence-preserving operation $\Lambda_k$, such that $\Lambda_k(|\phi\rangle)=|\psi\rangle$., Comment: 7 pages

The $k$-ME concurrence as a measure of multipartite entanglement (ME) unambiguously detects all $k$-nonseparable states in arbitrary dimensions, and satisfies many important properties of an entanglement measure. Negativity is a simple computable bipartite entanglement measure. Invariant and tangle are useful tools to study the properties of the quantum states. In this paper we mainly investigate the internal relations among the $k$-ME concurrence, negativity, polynomial invariants, and tangle. Strong links between $k$-ME concurrence and negativity as well as between $k$-ME concurrence and polynomial invariants are derived. We obtain the quantitative relation between $k$-ME ($k$=$n$) concurrence and negativity for all $n$-qubit states, give a exact value of the $n$-ME concurrence for the mixture of $n$-qubit GHZ states and white noise, and derive an connection between $k$-ME concurrence and tangle for $n$-qubit W state. Moreover, we find that for any $3$-qubit pure state the $k$-ME concurrence ($k$=2, 3) is related to negativity, tangle and polynomial invariants, while for $4$-qubit states the relations between $k$-ME concurrence (for $k$=2, 4) and negativity, and between $k$-ME concurrence and polynomial invariants also exist. Our work provides clear quantitative connections between $k$-ME concurrence and negativity, and between $k$-ME concurrence and polynomial invariants.

Uncertainty principle is the basis of quantum mechanics. It reflects the basic law of the movement of microscopic particles. Wigner-Yanase skew information, as a measure of quantum uncertainties, is used to characterize the intrinsic features of the state and the observable. In this paper, we mainly investigate the sum uncertainty relations for both quantum mechanical observables and quantum channels based on skew information. We establish a new uncertainty relation in terms of Wigner-Yanase skew information for $n$ observables, which is saturated (thus it holds as equality) for two incompatible observables. We also present two uncertainty relations for arbitrary finite $N$ quantum channels by using skew information. Our uncertainty relations have tighter lower bounds than the existing ones. Detailed examples are provided., Comment: 10 pages, 5 figures