Efficient learning of Scale-Adaptive Nearly Affine Invariant Networks.

Recent research has demonstrated the significance of incorporating invariance into neural networks. However, existing methods require direct sampling over the entire transformation set, notably computationally taxing for large groups like the affine group. In this study, we propose a more efficient approach by addressing the invariances of the subgroups within a larger group. For tackling affine invariance, we split it into the Euclidean group E(n) and uni-axial scaling group US(n), handling invariance individually. We employ an E(n)-invariant model for E(n)-invariance and average model outputs over data augmented from a US(n) distribution for US(n)-invariance. Our method maintains a favorable computational complexity of O(N2) in 2D and O(N4) in 3D scenarios, in contrast to the O(N6) (2D) and O(N12) (3D) complexities of averaged models. Crucially, the scale range for augmentation adapts during training to avoid excessive scale invariance. This is the first time nearly exact affine invariance is incorporated into neural networks without directly sampling the entire group. Extensive experiments unequivocally confirm its superiority, achieving new state-of-the-art results in affNIST and SIM2MNIST classifications while consuming less than 15% of inference time and fewer computational resources and model parameters compared to averaged models.

The effect of screw orientation on internal fixation of Letenneur type II Hoffa fractures: a biomechanics study.

BACKGROUND: To investigate the biomechanical effects of screw orientation and fracture block size on the internal fixation system for Letenneur type II Hoffa fractures. METHODS: The fracture models were randomly divided into six groups according to the fracture subtypes and the direction of nail placement, and a plumb line of the posterior condylar tangent was made across the base of the posterior femoral condyle. The fracture blocks of the three types of fracture were calculated and recorded in the sagittal position, and the biomechanical performance of the six groups was evaluated by biomechanical tests. The axial load on the fracture block at a displacement of 2 mm was set as the failure load, a gradually increasing axial load was applied to each fracture model using a customized indenter at a load of 250-750 N, and the displacements and failure loads of the six groups were recorded at different axial loads. RESULTS: Biomechanical test results showed that the larger the fracture block, the greater was the stability when nailing from front to back, and the smaller the fracture block, the greater was the strength when nailing from back to front (p < 0.001). As the fracture block became larger, the biomechanical advantage of nailing from posterior to anterior decreased.The displacement under 250 N load were 1.351 ± 0.113 mm, 1.465 ± 0.073 mm for Group IIa AP and Group IIa PA. The displacement under 500 N load were 2.596 ± 0.125 mm, 2.344 ± 0.099 mm for Group IIa AP and Group IIa PA. The displacement under 750 N load were 3.997 ± 0.164, 3.386 ± 0.125 mm for Group IIa AP and Group IIa PA. The failure loads were 384 ± 14 N, 415 ± 19 N for Group IIa AP and Group IIa PA. In the type IIa fracture group, the difference was no longer significant (p > 0.001). Therefore, there is a mechanical threshold that ranges from 38.36 to 52.33% between type IIa and type IIb fractures. CONCLUSIONS: The effect of the nailing direction on the strength of fixation has a fracture-block critical point, which is consistent overall with the trend that the larger the fracture block is, the greater the stability when nailing from anterior to posterior, and the smaller the fracture block is, the greater the strength when nailing from posterior to anterior.

[Research progress on biomechanics for internal fixation in tibial plateau fracture].

Objective: To review the biomechanical research progress of internal fixation of tibial plateau fracture in recent years and provide a reference for the selection of internal fixation in clinic. Methods: The literature related to the biomechanical research of internal fixation of tibial plateau fracture at home and abroad was extensively reviewed, and the biomechanical characteristics of the internal fixation mode and position as well as the biomechanical characteristics of different internal fixators, such as screws, plates, and intramedullary nails were summarized and analyzed. Results: Tibial plateau fracture is one of the common types of knee fractures. The conventional surgical treatment for tibial plateau fracture is open or closed reduction and internal fixation, which requires anatomical reduction and strong fixation. Anatomical reduction can restore the normal shape of the knee joint; strong fixation provides good biomechanical stability, so that the patient can have early functional exercise, restore knee mobility as early as possible, and avoid knee stiffness. Different internal fixators have their own biomechanical strengths and characteristics. The screw fixation has the advantage of being minimally invasive, but the fixation strength is limited, and it is mostly applied to Schatzker typeⅠfracture. For Schatzker Ⅰ-Ⅳ fracture, unilateral plate fixation can be used; for Schatzker Ⅴand Ⅵ fracture, bilateral plates fixation can be used to provide stronger fixation strength and avoid the stress concentration. The intramedullary nails fixation has the advantages of less trauma and less influence on the blood flow of the fracture end, but the fixation strength of the medial and lateral plateau is limited; so it is more suitable for tibial plateau fracture that involves only the metaphysis. Choosing the most appropriate internal fixation according to the patient's condition is still a major difficulty in the surgical treatment of tibial plateau fractures. Conclusion: Each internal fixator has good fixation effect on tibial plateau fracture within the applicable range, and it is an important research direction to improve and innovate the existing internal fixator from various aspects, such as manufacturing process, material, and morphology.

Numerical Study on the Unstable Flow Dynamics of Wormlike Micellar Solutions past a Sphere in the Creeping Flow Regime.

The flow dynamics of wormlike micellar solutions around a sphere is a fundamental problem in particle-laden complex fluids but is still understood insufficiently. In this study, the flows of the wormlike micellar solution past a sphere in the creeping flow regime are investigated numerically with the two species, micelles scission/reforming, Vasquez-Cook-McKinley (VCM) and the single-species Giesekus constitutive equations. The two constitutive models both exhibit the shear thinning and the extension hardening rheological properties. There exists a region with a high velocity that exceeds the main stream velocity in the wake of the sphere, forming a stretched wake with a large velocity gradient, when the fluids flow past a sphere at very low Reynolds numbers. We found a quasi-periodic fluctuation of the velocity with the time in the wake of the sphere using the Giesekus model, which shows a qualitative similarity with the results found in present and previous numerical simulations with the VCM model. The results indicate that it is the elasticity of the fluid that causes the flow instability at low Reynolds numbers, and the increase in the elasticity enhances the chaos of the velocity fluctuation. This elastic-induced instability might be the reason for the oscillating falling behaviors of a sphere in wormlike micellar solutions in prior experiments.

Tissue factor pathway inhibitor 2 suppresses the growth of thyroid cancer cells through by induction of apoptosis.

BACKGROUND: Tissue factor pathway inhibitor 2 (TFPI-2) has been recently identified as a tumor suppressor gene in several human cancers, whereas its role in thyroid cancer has been unclear. METHODS: The TFPI-2 protein level in thyroid cancer tissues and cell lines (8305C and B-CPAP) were examined using immunohistochemistry and immunoblotting. The TFPI-2 promoter methylation was examined using methylation-specific polymerase chain reaction (MSP). Lentivirus containing TFPI-2 cDNA (Lenti-TFPI-2) was constructed to elevate TFPI-2 expression in 8305C and B-CPAP cells. The effects of Lenti-TFPI-2 on cell proliferation in vitro and in vivo were evaluated by MTT assay and mouse xenograft model. Annexin V/PI double staining assay was performed to detect the effect of Lenti-TFPI-2 on cell apoptosis. RESULTS: TFPI-2 protein level were decreased in cancer tissues and lymph node metastasis, and TFPI-2 protein level is positively associated with survival time. The promoter of TFPI-2 is hypermethylated in cancer tissues. TFPI-2 mRNA and protein levels were abundant in normal human thyroid follicular cell line Nthy-ori 3-1 cells, whereas they were decreased in 8305C and B-CPAP cells. pcDNA-TFPI-2 elevated TFPI-2 mRNA and protein in 8305C and B-CPAP cells. TFPI-2 overexpression suppressed proliferation and induced apoptosis of 8305C and B-CPAP cells. CONCLUSIONS: TFPI-2 inactivation may play a role in thyroid cancer tumorigenesis and development. TFPI-2 overexpression suppressed cell proliferation through induction of cell apoptosis, suggesting that TFPI-2 may serve as a novel and effective target for thyroid cancer therapy.

Material Synthesis and Device Aspects of Monolayer Tungsten Diselenide.

In this paper, we investigate the synthesis of WSe2 by chemical vapor deposition and study the current transport and device scaling of monolayer WSe2. We found that the device characteristics of the back-gated WSe2 transistors with thick oxides are very sensitive to the applied drain bias, especially for transistors in the sub-micrometer regime. The threshold voltage, subthreshold swing, and extracted field-effect mobility vary with the applied drain bias. The output characteristics in the long-channel transistors show ohmic-like behavior, while that in the short-channel transistors show Schottky-like behavior. Our investigation reveals that these phenomena are caused by the drain-induced barrier lowering (short-channel effect). For back-gated WSe2 transistors with 280 nm oxide, the short-channel effect appears when the channel length is shorter than 0.4 µm. This extremely long electrostatic scaling length is due to the thick back-gate oxides. In addition, we also found that the hydrogen flow rate and the amount of WO3 precursor play an important role in the morphology of the WSe2. The hole mobility of the monolayer WSe2 is limited by Columbic scattering below 250 K, while it is limited by phonon scattering above 250 K. These findings are very important for the synthesis of WSe2 and accurate characterization of the electronic devices based on 2D materials.

Thermal sublimation: a scalable and controllable thinning method for the fabrication of few-layer black phosphorus.

We report uniform layer-by-layer sublimation of black phosphorus under heating below 600 K. The uniformity and crystallinity of BP samples after thermal thinning were confirmed by Raman spectra and Raman mapping. The sublimation rate of BP was around 0.18 nm min-1 at 500 K and 1.15 nm min-1 at 550 K. Both room and high temperature Raman peak intensity ratio [Formula: see text] as functions of BP thickness were established for in situ thickness determination and control. Uniform and crystalline 2 to 4-layer BP flakes with areas from 10 to 1000 µm2 were prepared with this method. No micron scale defects were observed. The sublimation thinning method was shown to be a controllable and scalable approach to prepare high-quality few-layer black phosphorus.

Selective and sensitive fluorescence-shift probes based on two dansyl groups for mercury(ii) ion detection.

Two probes ( and ) bearing two dansyl fluorophores were synthesized and applied to the detection of mercury(ii) ions in aqueous solution. These probes exhibited a selective response to Hg(2+) in a buffered solution, with high sensitivity and a unique fluorescence response signal which displayed a blue-shift effect in the fluorescence emission peak. The Hg(2+) recognition mechanisms of the probes were determined by NMR spectroscopy, ESI-MS and UV-vis spectroscopy. The results showed that probe and mercury(ii) ions formed an unusual 2:2 stoichiometric ratio complex, while probe and Hg(2+) formed a multidentate complex with a stoichiometric ratio of 2:1.

Bis(µ-2-hydroxy-benozato)-κO,O':O';κO:O,O'-bis-[(2-hydroxy-benozato-κO,O')(1,10-phenanthroline-κN,N')cadmium(II)].

The dinuclear title compound, [Cd(2)(C(7)H(5)O(3))(4)(C(12)H(8)N(2))(2)], is located on a crystallographic rotation twofold axis. The two Cd(II) ions are connected by two tridentate bridging 2-hydroxy-benzoate anions. Each Cd(II) ion is seven-coordinated by five O atoms from three 2-hydroxy-benzoate ligands and two N atoms from 1,10-phenanthroline. The 2-hydroxy-benzoate mol-ecules adopt two kinds of coordination mode, bidentate chelating and tridentate bridging-chelating. Intra-molecular hydrogen bonds between hydr-oxy and carboxyl-ate groups from 2-hydroxy-benzoate groups and π-π stacking interactions between parallel 1,10-phenanthroline ligands [centroid-centroid distances = 3.707 (3) and 3.842 (3) Å] are observed. Furthermore, adjacent benzene rings from 2-hydroxy-benzoate ligands are involved in π-π inter-actions with inter-planar distances of 3.642 (3) Å, thereby forming a chain along the a axis direction.