Introduction: Late-life depression (LLD) is associated with cognitive deficit with risk of future dementia. By examining the entropy of the spontaneous brain activity, we aimed to understand the neural mechanism pertaining to cognitive decline in LLD.

Methods: We collected MRI scans in older adults with LLD (n = 32), mild cognitive impairment [MCI (n = 25)] and normal cognitive function [NC, (n = 47)]. Multiscale entropy analysis (MSE) was applied to resting-state fMRI data. Under the scale factor (tau) 1 and 2, reliable separation of fMRI data and noise was achieved. We calculated the brain entropy in 90 brain regions based on automated anatomical atlas (AAL). Due to exploratory nature of this study, we presented data of group-wise comparison in brain entropy between LLD vs. NC, MCI vs. NC, and LLD and MCD with a p-value below 0.001.

Results: The mean Mini-Mental State Examination (MMSE) score of LLD and MCI was 27.9 and 25.6. Under tau 2, we found higher brain entropy of LLD in left globus pallidus than MCI (p = 0.002) and NC (p = 0,009). Higher brain entropy of LLD than NC was also found in left frontal superior gyrus, left middle superior gyrus, left amygdala and left inferior parietal gyrus. The only brain region with higher brain entropy in MCI than control was left posterior cingulum (p-value = 0.015). Under tau 1, higher brain entropy was also found in LLD than in MCI in right orbital part of medial frontal gyrus and left globus pallidus (p-value = 0.007 and 0.005).

Conclusions: Our result is consistent with prior hypothesis where higher brain entropy was found during early aging process as compensation. We found such phenomenon particular in left globus pallidus in LLD, which could be served as a discriminative brain region. Being a key region in reward system, we hypothesis such region may be associated with apathy and with unique pathway of cognitive decline in LLD. We will undertake subsequent analysis longitudinally in this cohort

Key words: resting-state fMRI, Late-life depression, Brain entropy, globus pallidus, cognitive decline

The delay-shift of the pre-pulse may mislead the determination of its origination and cause problems for the temporal contrast improvement of high-peak-power lasers, especially when the corresponding post-pulse is beyond the time window of the measurement device. In this work, an empirical formula is proposed to predict the delay-shift of pre-pulses for the first time. The empirical formula shows that the delay-shift is proportional to the square of the post-pulse’s initial delay, and also the ratio of the third-order dispersion to the group delay dispersion’s square, which intuitively reveals the main cause for the delay-shift and may provide a convenient routing for identifying the real sources of pre-pulses in both chirped-pulse amplification (CPA) and optical parametric chirped-pulse amplification (OPCPA) systems. The empirical formula agrees well with the experimental results both in the CPA and the OPCPA systems. Besides, a numerical simulation is also carried out to further verify the empirical formula.

To meet the demands of laser-ion acceleration at a high repetition rate, we have developed a comprehensive diagnostic system for real-time and in situ monitoring of liquid sheet targets (LSTs). The spatially resolved rapid characterizations of an LST’s thickness, flatness, tilt angle and position are fulfilled by different subsystems with high accuracy. With the help of the diagnostic system, we reveal the dependence of thickness distribution on collision parameters and report the 238-nm liquid sheet generated by the collision of two liquid jets. Control methods for the flatness and tilt angle of LSTs have also been provided, which are essential for applications of laser-driven ion acceleration and others.

Post-acceleration of protons in helical coil targets driven by intense, ultrashort laser pulses can enhance ion energy by utilizing the transient current from the targets’ self-discharge. The acceleration length of protons can exceed a few millimeters, and the acceleration gradient is of the order of GeV/m. How to ensure the synchronization between the accelerating electric field and the protons is a crucial problem for efficient post-acceleration. In this paper, we study how the electric field mismatch induced by current dispersion affects the synchronous acceleration of protons. We propose a scheme using a two-stage helical coil to control the current dispersion. With optimized parameters, the energy gain of protons is increased by four times. Proton energy is expected to reach 45 MeV using a hundreds-of-terawatts laser, or more than 100 MeV using a petawatt laser, by controlling the current dispersion.

The mitochondrial genome provides important information for phylogenetic analysis and an understanding of evolutionary origin. In this study, the mitochondrial genomes of Ilisha elongata and Setipinna tenuifilis were sequenced, which are typical circular vertebrate mitochondrial genomes composed of 16,770 and 16,805 bp, respectively. The mitogenomes of I. elongata and S. tenuifilis include 13 protein-coding genes (PCGs), 22 transfer RNA (tRNA), two ribosomal RNA (rRNA) genes and one control region (CR). Both two species' genome compositions were highly A + T biased and exhibited positive AT-skews and negative GC-skews. The genetic distance and Ka/Ks ratio analyses indicated that 13 PCGs were affected by purifying selection and the selection pressures were different from certain deep-sea fishes, which were most likely due to the difference in their living environment. Results of phylogenetic analysis support close relationships among Chirocentridae, Denticipitidae, Clupeidae, Engraulidae and Pristigasteridae based on the nucleotide and amino acid sequences of 13 PCGs. Within Clupeoidei, I. elongata and S. tenuifilis were most closely related to the family Pristigasteridae and Engraulidae, respectively. These results will help to better understand the evolutionary position of Clupeiformes and provide a reference for further phylogenetic research on Clupeiformes species.

This paper describes gft (general fine-tuning), a little language for deep nets, introduced at an ACL-2022 tutorial. gft makes deep nets accessible to a broad audience including non-programmers. It is standard practice in many fields to use statistics packages such as R. One should not need to know how to program in order to fit a regression or classification model and to use the model to make predictions for novel inputs. With gft, fine-tuning and inference are similar to fit and predict in regression and classification. gft demystifies deep nets; no one would suggest that regression-like methods are “intelligent.”

A droplet impinging on a superhydrophobic substrate in an electric field is an important process in droplet manipulation and electrostatic spraying. Here, the entire impinging dynamic of the droplet in a vertical electric field is studied by a visualization experiment and numerical simulation with OpenFOAM. We investigate the effect of an electrostatic force on droplet impact in depth, where four ejection modes and three rebound modes are found experimentally. In particular, the filamentous ejecting phenomenon occurs after a droplet impinging on a superhydrophobic substrate is first discovered. In the numerical simulation, the strong coupling between the dynamic distribution of the interface electric charge and the evolution of the droplet profile can lead to different ejection modes, and the different ejection behaviours are caused by the combined effects of electrostatic pressure, capillary pressure, dynamic pressure and static pressure on the droplet apex. A charge scaling law for the ejection droplets is proposed. Furthermore, a set of theoretical models is established, which can successfully predict the threshold electric capillary number for different droplet ejection modes. The results reveal some important characteristics for a droplet impinging on a superhydrophobic surface in an electric field, which could facilitate the design of electrically operated droplet equipment and guide the safe and stable operation of the device.

This paper considers a generalized panel data transformation model with fixed effects where the structural function is assumed to be additive. In our model, no parametric assumptions are imposed on the transformation function, the structural function, or the distribution of the idiosyncratic error term. The model is widely applicable and includes many popular panel data models as special cases. We propose a kernel-based nonparametric estimator for the structural function. The estimator has a closed-form solution and is easy to implement. We study the asymptotic properties of our estimator and show that it is asymptotically normally distributed. The Monte Carlo simulations demonstrate that our new estimator performs well in finite samples.

The weakest possible waves in nature are detectable by improving sensitive measurements, but the attainable imaging resolution of low-frequency waves is still challenging, especially in aeroacoustic experiments. In this work, we show how extended-resolution imaging of low-frequency wave sources can be achieved by incorporating acoustic analogy into tomography. First, an equivalent source of sound, which is dependent on the low-frequency target sound field, is produced due to the nonlinear coupling and interaction with an external high-frequency incident plane wave. Next, the low-frequency sources are reconstructed based on the induced sound waves recorded at the receivers. The induced sound waves are of high frequency to enable the extended-resolution imaging. The physical processes involved are theoretically explained based on the insightful acoustic analogy theory and the Born approximation. The numerical and experimental demonstration cases, with representative but straightforward configurations, show that the proposed method can identify the isolated target sources (at low frequencies) with a separation distance smaller than one-tenth to one-thirtieth of the wavelength, yielding much better resolution than the conventional acoustic imaging approaches. The results suggest that the proposed method will be a promising candidate to investigate the properties of an acoustic source within small regions, and, therefore, likely to be used in the study of the associated fluid physics.

The time to positivity (TTP) of blood cultures has been considered a predictor of clinical outcomes for bacteremia. This retrospective study aimed to determine the clinical value of TTP for the prognostic assessment of patients with Escherichia coli bacteremia. A total of 167 adult patients with E.coli bacteremia identified over a 22-month period in a 3500-bed university teaching hospital in China were studied. The standard cut-off TTP was 11 h in the patient cohort. The septic shock occurred in 27.9% of patients with early TTP (⩽11 h) and in 7.1% of those with a prolonged TTP (>11 h) (P = 0.003). The mortality rate was significantly higher for patients in the early than in the late group (17.7% vs. 4.0%, P < 0.001). Multivariate analysis showed that an early TTP (OR 4.50, 95% CI 1.70–11.93), intensive care unit admission (OR 8.39, 95% CI 2.01–35.14) and neutropenia (OR 4.20, 95% CI 1.55–11.40) were independently associated with septic shock. Likewise, the independent risk factors for mortality of patients were an early TTP (OR 3.80, 95% CI 1.04–12.90), intensive care unit admission (OR 6.45; 95% CI 1.14–36.53), a Pittsburgh bacteremia score ⩾2 (OR 4.34, 95% CI 1.22–15.47) and a Charlson Comorbidity Index ⩾3 (OR 11.29, 95% CI 2.81–45.39). Overall, a TTP for blood cultures within 11 h appears to be associated with worse outcomes for patients with E.coli bacteremia.

Glacial retreat in response to warming climates in the arid Xinjiang region of northwestern China directly impacts downstream water resources available for local communities. We used high-resolution satellite imagery from 1969 to 2014 to delineate spatial changes in 54 active glaciers in the upper Kaidu River Basin in the Tian Shan as well as their past expanses during the Little Ice Age (LIA). We manually delineated their boundaries based on the interpretation of glacial, geomorphic and topographic features. From the total glacier surface area, we estimated glacier volume and mass. From 1969 to 2014, glacier area decreased by 10.1 ± 1.0 km2 (relative loss of 34.2 ± 3.5%) and mass by 1.025 ± 0.108 Gt (relative loss of 43 ± 4.6%). From the LIA maximum (est. 1586 CE) to 1969, relative losses were less (25.7 ± 4.3% area loss and 33.1 ± 5.7% mass loss). Our results indicate that glacier recession is accelerating over time and that the glaciers are currently losing over 1.5 times more relative area than elsewhere in the Tian Shan. Using linear and non-linear projections, we estimate that these glaciers may disappear between 2050 and 2150 CE if climatic warming continues at the same pace.

Activated carbon (AC) has been widely used as catalyst for oxygen reduction reaction (ORR) in air-cathode microbial fuel cells (MFCs). Here we demonstrate a new method to improve the AC air-cathode by blending it with reduced graphene oxide (rGO). rGO sheets are first deposited on Ni foam and AC is then brushed onto it with controlled mass loading. rGO sheets not only improve the electrical conductivity of AC, but also provide a large number of ORR areas. Rotating ring disk electrode measurements reveal that the number of transferred electrons at rGO-AC cathode is 3.5, indicating the four-electron pathway is the dominant process. Significantly, the MFC with rGO-AC cathode delivers a maximum power density of 2.25 ± 0.05 W/m2, which is substantially higher than that of plain AC cathode (1.35 ± 0.07 W/m2) and those for other air-cathode MFCs using AC as ORR catalyst under the same mass loading.