Search

Purpose: The central nervous mechanism of acute tinnitus is different from that of chronic tinnitus, which may be related to the difference of cerebral blood flow (CBF) perfusion in certain regions. To verify this conjecture, we used arterial spin labeling (ASL) perfusion magnetic resonance imaging (MRI) in this study to compare the CBF alterations of patients with acute and chronic tinnitus.Methods: The current study included patients with chronic tinnitus (n = 35), acute tinnitus (n = 30), and healthy controls (n = 40) who were age-, sex-, and education-matched. All participants underwent MRI scanning and then ASL images were obtained to measure CBF of the entire brain and analyze the differences between groups as well as the correlations with tinnitus characteristics.Results: The chronic tinnitus group showed increased z-CBF in the right superior temporal gyrus (STG) and superior frontal gyrus (SFG) when compared with the acute tinnitus patients. Further connectivity analysis found enhanced CBF connectivity between the right STG and fusiform gyrus (FG), the right SFG and left middle occipital gyrus (MOG), as well as the right parahippocampal gyrus (PHG). Moreover, in the chronic tinnitus group, the tinnitus handicap questionnaire (THQ) score was positively correlated with the normalized z-CBF of right STG (r = 0.440, p = 0.013).Conclusion: Our results confirmed that the CBF changes in some brain regions were different between acute and chronic tinnitus patients, which was correlated with certain tinnitus characteristics. This is of great value to further research on chronicity of tinnitus, and ASL has a promising application in the measurement of CBF.

In this paper, a non-autonomous memristive FitzHugh-Nagumo (FHN) circuit is constructed using a second-order memristive diode bridge with LC network. For convenience of circuit implementation, an AC voltage source is adopted to substitute the original AC current stimulus. Stimulated by the slowly varying AC voltage source, the number, locations and stabilities of the equilibrium points slowly evolve with the time, which are thus indicated as the AC equilibrium points. Different sequences of fold and/or Hopf bifurcations are encountered in a full period of time series evolutions, leading to various kinds of chaotic or periodic bursting activities. To figure out the related bifurcation mechanisms, the fold and Hopf bifurcation sets are mathematically formulated to locate the critical bifurcation points. On this basis, the transitions between the resting and repetitive spiking states are clearly illustrated by the time series of the AC equilibrium points and state variables, from which Hopf/subHopf, Hopf/Hopf, and Hopf/fold bursting oscillations are identified in the specified parameter regions. Finally, based on a fabricated hardware circuit, the experimental measurements are executed. The results verify that the presented memristive FHN circuit indeed exhibits complex bursting activities, which enriches the family of memristor-based FHN circuits with bursting dynamics.

To address the problem of “pseudochange” caused by illumination, phasing, and shadows in multiperiod remote sensing images, a bitemporal, hyperspectral remote sensing image change detection method is proposed. The method uses a pyramid self-attention mechanism, and the attention module is introduced to simulate the attention mechanism of human eyes, where more attention is given to a small number of important objects. The model architecture uses a general encoder–decoder paradigm, in which shared encoders extract common features, whereas individual decoders learn task-specific representations. The classifier takes fused data to locate where the changes occurred. The method is tested on the ZY1-CD dataset.

As wind energy is a kind of clean energy, wind power generation industry has developed rapidly in recent years. Therefore, it is of great significance for the power industry and relevant regulatory authorities to obtain the number and spatial distribution of wind turbine quickly based on remote sensing images. Despite of the rapid development of remote sensing information extraction, the traditional algorithms based on remote sensing classification and recognition cannot meet the requirements of wind turbine extraction on a large-scale area due to the insufficient feature expression of remote sensing images. In this paper, an iterative training method based on the Faster R-CNN (the target extract framework of deep learning) which uses the positive and negative sample set obtained from the first target detection result of ResNet classification for training again has been proposed, which has got the 97.5 percent accuracy for wind turbine target extraction on the 2m spatial resolution map of domestic satellite remote sensing image on the whole china.

Tinnitus refers to sound perception in the absence of external sound stimulus. It has become a worldwide problem affecting all age groups especially the elderly. Tinnitus often accompanies hearing loss and some mood disorders like depression and anxiety. The comprehensive adverse effects of tinnitus on people determine the severity of tinnitus. Understanding the mechanisms of tinnitus and related discomfort may be beneficial to the prevention and treatment, and then getting patients out of tinnitus distress. Functional magnetic resonance imaging (fMRI) is a powerful technique for characterizing the intrinsic brain activity and making us better understand the tinnitus neural mechanism. In this article, we review fMRI studies published in recent years on the neuroimaging mechanisms of tinnitus. The results have revealed various neural network alterations in tinnitus patients, including the auditory system, limbic system, default mode network, attention system, and some other areas involved in memory, emotion, attention, and control. Moreover, changes in functional connectivity and neural activity in these networks are related to the perception, persistence, and severity of tinnitus. In summary, the neural mechanism of tinnitus is a complex regulatory mechanism involving multiple networks. Future research is needed to study these neural networks more accurately to refine the tinnitus models.

Cloud detection is one of the most important tasks and key technologies during remote sensing image processing. This paper presented a cloud detection method based on multitemporal thermal infrared data. Compared with traditional methods, this method could not only provide a better discrimination of cloud-free and cloud-cover pixels, but also have the anti-interference ability for snow. Landsat8 data around Beijing and Hotan district were selected for checking the accuracy of this method. The validation results showed that this method could extract cloud boundary accurately, and had a good extraction effect for thin clouds. Most important of all, it could accurately distinguish cloud pixels from snow cover. Whether in flat areas or mountain areas, for thick clouds or scattered thin clouds, the overall accuracy of this method could be always kept above 95%. This showed that the proposed method would have a very broad application prospect in cloud detection of remote sensing images.

The difference of multi-label classification from traditional classification is that an instance may associate a set of labels simultaneously. In recent study, some scholars have proposed that the information which derives from the query instance’s nearest neighbors, can be useful when predicting the labels of the query instance. On the basis of their research, we propose a new approach to multi-label classification, which employs neural tensor network (NTN) to explore the relations among the labels of neighbors and classify the query instance with these correlations. This method utilizes the correlations and interdependencies between labels and leverages the potential of data. Experiments on real data show that our method can achieve good performance in multi-label classification.

Background: Chronic tinnitus is often accompanied with headache symptom that will affect the cerebral blood flow (CBF) and exacerbate the tinnitus distress. However, the potential relationship between headache and tinnitus remains unclear. This study will investigate whether aberrant CBF patterns exist in chronic tinnitus patients and examine the influence of headache on CBF alterations in chronic tinnitus. Methods: Participants included chronic tinnitus patients (n=45) and non-tinnitus controls (n=50), matched for age, sex, education, and hearing thresholds. CBF images were collected and analyzed using arterial spin labeling (ASL) perfusion functional magnetic resonance imaging (fMRI). Regions with major CBF differences between tinnitus patients and non-tinnitus controls were first detected. The interaction effects between headache and tinnitus for CBF alterations were further examined. Correlation analyses illustrated the association between CBF values and tinnitus severity as well as between CBF and degree of headache.Results: Compared with non-tinnitus controls, chronic tinnitus patients exhibited decreased CBF, primarily in right superior temporal gyrus (STG), bilateral middle frontal gyrus (MFG), and left superior frontal gyrus (SFG); decreased CBF in these regions was correlated with tinnitus distress. There was a significant interaction effect between headache and tinnitus for CBF in right STG and MFG. Moreover, the degree of headache correlated negatively with CBF in tinnitus patients.Conclusions: Chronic tinnitus patients exhibited reduced CBF in the auditory and prefrontal cortex. Headache may facilitate a CBF decrease in the setting of tinnitus, which may underlie the neuropathological mechanisms of chronic tinnitus comorbid with headache.

There are two main challenges in the diagnosis of blood cancer. The first is to diagnose cancer from healthy control, and the second is to identify the types of blood cancer. The chemometrics method combined with laser-induced breakdown spectroscopy (LIBS) can be used for cancer detection. However, chemometrics methods were easily influenced by the spectral feature redundancy and noise, resulting in low accuracy rate because of their simple structure. We proposed an approach using LIBS combined with the ensemble learning based on the random subspace method (RSM). The serum samples were dripped onto a boric acid substrate for LIBS spectrum collection. The complete blood cancer sample set include leukemia [acute myeloid leukemia (AML) and chronic myelogenous leukemia (CML)], multiple myeloma (MM), and lymphoma. The results showed that the accuracy rates using k nearest neighbors (kNN) and linear discriminant analysis (LDA) only were 88.14% and 94.45%, respectively, while using RSM with LDA (RSM-LDA), the average accuracy rate was improved from 94.45% to 98.34%. Furthermore, the variable importance of spectral lines (Na, K, Mg, Ca, H, O, N, C-N) were evaluated by the RSM-LDA model, which can improve the recognition ability of blood cancer types. Comparing the RSM-LDA model and only with LDA, the results showed that the average accuracy rate for cancer type identification was improved from 80.4% to 91.0%. These results demonstrate that LIBS combined with the RSM-LDA model can discriminate the blood cancer from the health control, as well as the recognition the types for blood cancers.

With the rapid development of Internet-related industries, the Internet news is containing more and more information as well as being paid more attention by human beings. In order to explore the intrinsic links among the news and extracted effective topics, we put forward ProjectRank model based on latent Dirichlet allocation. Large-scale Chinese text sets serves as initial input of this model. By combining LDA topic model with distance between probability distributions, we figure out scores of topics so as to quantify the result of topic evaluation. Moreover, this model proves to be useful to track trends of hot news in different time periods. This paper applies ProjectRank model to real Internet news data obtained from several websites and achieves good experiment results.

To make better use of the advantages of radar and promote the application of image fusion based on radar data, the author uses different fusion methods based on GF-3 SAR and GF-1 MSS, and evaluates the fusion results by analyzing mean, variance, information entropy, average gradient, spectral distortion and correlation coefficient. The results show that HSV and GS transforms have the best performances in overall. PC is recognized as the third, while it is still remarkable that it has the best ability of spectral retention. And the specialty in NIR band makes PC more conducive for extraction of vegetation. Brovey and Multiplicative transforms are not effective in comparison.

Topic evolution analysis is an important task in text mining to help people understand complex relationships between latent topics in large text collections and discover interesting topic evolution patterns. As most of the studies of topic evolution are made on open data set and few achievements are gained on textual visualization in the domain of finance, we propose a visual topic model named FinancialFlow to explore topic evolution patterns from financial news. Our model can be described as an integration of topic modeling algorithms and text visualization techniques, the key of which is to employ the hierarchical Dirichlet process to discover coherent topics in text data with temporal information and then visualize the process of how topics emerge, decline and develop in the form of flows. Finally, model evaluation and experimental results verify the feasibility and effectiveness of our proposed method in the real-world financial news data.

Ta2O5 films deposited by ion beam sputtering were post-treated with different methods. The optical properties, microstructures, surface morphologies, absorption and laser-induced damage threshold (LIDT) at 1064 nm and 12 ns of the films were comparatively studied. Annealing conduced to larger changes of optical transmittance and surface morphologies whereas it led to smaller changes of absorption and LIDT than laser conditioning. As it was shown, all the three post-treatment methods in this study were beneficial to the laser damage resistance of the films, yet the details of the defect elimination caused by these methods were different. A defect evolution model was proposed, which revealed the different effects of post-treatment methods on the defect repair of the films. The highest LIDT achieved by the combination of annealing and laser conditioning was due to their complementary effect which resulted in the lowest defect density and the smallest defect size of the film.