WOOD: Wasserstein-Based Out-of-Distribution Detection.

The training and testing data for deep-neural-network-based classifiers are usually assumed to be sampled from the same distribution. When part of the testing samples are drawn from a distribution that is sufficiently far away from that of the training samples (a.k.a. out-of-distribution (OOD) samples), the trained neural network has a tendency to make high-confidence predictions for these OOD samples. Detection of the OOD samples is critical when training a neural network used for image classification, object detection, etc. It can enhance the classifier's robustness to irrelevant inputs, and improve the system's resilience and security under different forms of attacks. Detection of OOD samples has three main challenges: (i) the proposed OOD detection method should be compatible with various architectures of classifiers (e.g., DenseNet, ResNet) without significantly increasing the model complexity and requirements on computational resources; (ii) the OOD samples may come from multiple distributions, whose class labels are commonly unavailable; (iii) a score function needs to be defined to effectively separate OOD samples from in-distribution (InD) samples. To overcome these challenges, we propose a Wasserstein-based out-of-distribution detection (WOOD) method. The basic idea is to define a Wasserstein-based score that evaluates the dissimilarity between a test sample and the distribution of InD samples. An optimization problem is then formulated and solved based on the proposed score function. The statistical learning bound of the proposed method is investigated to guarantee that the loss value achieved by the empirical optimizer approximates the global optimum. The comparison study results demonstrate that the proposed WOOD consistently outperforms other existing OOD detection methods.

Free dermofat grafting for chest deformity in microtia reconstruction.

OBJECTIVE: Chest deformity is one of the complications that occurs after costal cartilage harvesting for auricle reconstruction. In this study, we presented a novel method of free dermofat grafting to repair cartilage defect and aimed to evaluate its effect in ameliorating chest deformity. METHODS: Seventy-six pediatric patients were included in the study, comprising free dermofat grafting group (n = 38) and control group (n = 38). After harvesting costal cartilage, empty perichondrial space of right seventh costal cartilage was filled with free dermofat grafts in free dermofat grafting group. Thoracic computed tomography (CT) was performed three months after surgery and 3D colormap quantification was performed to quantify chest surface asymmetry. The quantified data were further analyzed to compare chest asymmetry level in free dermofat grafting and control groups. RESULTS: In the free dermofat grafting group, the mean of asymmetry level was 2.2 mm. While in the control group, the mean of asymmetry was 5.7 mm. After a comparison between the two groups, the level of asymmetry showed a significant difference (p < 0.01). CONCLUSION: Free dermofat grafting method is easy to perform and a feasible option for ameliorating chest deformity in microtia reconstruction.

OsEIL1 protects rice growth under NH4+ nutrition by regulating OsVTC1-3-dependent N-glycosylation and root NH4+ efflux.

Rice is known for its superior adaptation to ammonium (NH4+ ) as a nitrogen source. Compared to many other cereals, it displays lower NH4+ efflux in roots and higher nitrogen-use efficiency on NH4+ . A critical role for GDP-mannose pyrophosphorylase (VTC1) in controlling root NH4+ fluxes was previously documented in Arabidopsis, but the molecular pathways involved in regulating VTC1-dependent NH4+ efflux remain unclear. Here, we report that ETHYLENE-INSENSITIVE3-LIKE1 (OsEIL1) acts as a key transcription factor regulating OsVTC1-3-dependent NH4+ efflux and protein N-glycosylation in rice grown under NH4+ nutrition. We show that OsEIL1 in rice plays a contrasting role to Arabidopsis-homologous ETHYLENE-INSENSITIVE3 (AtEIN3) and maintains rice growth under NH4+ by stabilizing protein N-glycosylation and reducing root NH4+ efflux. OsEIL1 constrains NH4+ efflux by activation of OsVTC1-3, but not OsVTC1-1 or OsVTC1-8. OsEIL1 binds directly to the promoter EIN3-binding site (EBS) of OsVTC1-3 in vitro and in vivo and acts to increase the transcription of OsVTC1-3. Our work demonstrates an important link between excessive root NH4+ efflux and OsVTC1-3-mediated protein N-glycosylation in rice grown under NH4+ nutrition and identifies OsEIL1 as a direct genetic regulator of OsVTC1-3 expression.

Secondary surgery for the unsatisfactory auricle after auricular reconstruction.

OBJECTIVES: Complications, including framework exposure, infections, and reconstructed auricle deformation, may occur after auricular reconstruction. However, reports on surgical methods for cases with unsatisfactory outcomes after auricular reconstruction using an autologous costal cartilage are insufficient. Herein, we summarized retreatment casesfor poor ear morphology in patients who had undergone auricular reconstruction in our department for 5 years and discussed other techniques. METHODS: Between September 2014 and September 2019, 24 ears of 24 patients with poor morphology, unsatisfactory macroscopic characteristics and anatomical structures, and unsatisfactory outcomes of local repair after auricular reconstructive surgery were treated. Patients were divided into the following three groups: type 1 (9 ears), with intact and sufficient hairless skin in the mastoid region behind the reconstructed ear; type 2 (7 ears), with intact, but insufficient, hairless skin in the mastoid region behind the reconstructed ear; and type 3 (8 ears), with hairless skin in the mastoid region behind the reconstructed ear with impaired skin integrity. RESULTS: Twenty-two (91.6%) patients successfully completed the surgical treatment and recovered well; one experienced delayed wound healing and another developed hypertrophic scarring at the incision site at 3 months postoperatively. All patients were followed for 0.5-4 (mean, 2.8) years. The macrostructure of the reconstructed ear post-revision was stable and significantly improved in terms of morphology and structure. CONCLUSIONS: In patients with unsatisfactory outcomes after auricular reconstruction, the appropriate technique for the revision surgery should consider the local soft tissue conditions of the reconstructed ear to obtain satisfactory results.

A Shape-Constrained Neural Data Fusion Network for Health Index Construction and Residual Life Prediction.

With the rapid development of sensor technologies, multisensor signals are now readily available for health condition monitoring and remaining useful life (RUL) prediction. To fully utilize these signals for a better health condition assessment and RUL prediction, health indices are often constructed through various data fusion techniques. Nevertheless, most of the existing methods fuse signals linearly, which may not be sufficient to characterize the health status for RUL prediction. To address this issue and improve the predictability, this article proposes a novel nonlinear data fusion approach, namely, a shape-constrained neural data fusion network for health index construction. Especially, a neural network-based structure is employed, and a novel loss function is formulated by simultaneously considering the monotonicity and curvature of the constructed health index and its variability at the failure time. A tailored adaptive moment estimation algorithm (Adam) is proposed for model parameter estimation. The effectiveness of the proposed method is demonstrated and compared through a case study using the Commercial Modular Aero-Propulsion System Simulation (C-MAPSS) data set.

Study on a linear relationship between limited pressure difference and coil current of on/off valve and its influential factors.

On/off solenoid valves with PWM control are widely used in all types of vehicle electro-hydraulic control systems respecting to their desirable properties of reliable, low cost and fast acting. However, it can hardly achieve a linear hydraulic modulation by using on/off valves mainly due to the nonlinear behaviors of valve dynamics and fluid, which affects the control accuracy significantly. In this paper, a linear relationship between limited pressure difference and coil current of an on/off valve in its critical closed state is proposed and illustrated, which has a great potential to be applied to improve hydraulic control performance. The hydraulic braking system of case study is modeled. The linear correspondence between limited pressure difference and coil current of the inlet valve is simulated and further verified experimentally. Based on validated simulation models, the impacts of key parameters are researched. The limited pressure difference affected by environmental temperatures is experimentally studied, and the amended linear relation is given according to the test data.