Search

The dual neural network (DNN)-based $k$ -winner-take-all (WTA) model is able to identify the $k$ largest numbers from its $m$ input numbers. When there are imperfections, such as non-ideal step function and Gaussian input noise, in the realization, the model may not output the correct result. This brief analyzes the influence of the imperfections on the operational correctness of the model. Due to the imperfections, it is not efficient to use the original DNN- $k$ WTA dynamics for analyzing the influence. In this regard, this brief first derives an equivalent model to describe the dynamics of the model under the imperfections. From the equivalent model, we derive a sufficient condition for which the model outputs the correct result. Thus, we apply the sufficient condition to design an efficiently estimation method for the probability of the model outputting the correct result. Furthermore, for the inputs with uniform distribution, a closed form expression for the probability value is derived. Finally, we extend our analysis for handling non-Gaussian input noise. Simulation results are provided to validate our theoretical results.

The pronounced brittleness of hard Laves phase intermetallics is detrimental to their tribological properties at room temperature. In this study, we utilized a heterogeneous structure to engineer an ultrastrong dual-phase (Laves + B2) AlCoFeNiNb high-entropy alloy that exhibits a low wear rate (3.82×10−6mm3/(N·m)) at room temperature. This wear resistance in the ball-on-disc sliding friction test with the counterpart of Al2O3balls stems from the activated deformation ability in the ultrafine Laves lamellae under heterogeneous interface constraints. Furthermore, as tribological stress intensifies, the surface deformation mechanism transitions from dislocation slip on the basal and pyramidal planes to a unique combination of local shear and grain rotation within the Laves phase. Our study illuminates fresh perspectives for mitigating the embrittling effect of Laves phase intermetallics under tribological loading and for the development of wear-resistant materials.

The sigmoid function is a representative activation function in shallow neural networks. Its hardware realization is challenging due to the complex exponential and reciprocal operations. Existing studies applied piecewise models to approximate sigmoid function and employed numerical methods or non-uniform input segmentations to mitigate fitting inaccuracies. However, the breakpoints introduce inevitable approximation precision loss. Besides, additional fitting processes greatly increase hardware complexity and power consumption. This paper presents a hardware-friendly sigmoidal approximation from the perspective of probability theory. We find that for a given input, the output of a sigmoid function can be approximated by the probability that the sum of this input and a Gaussian random variable is greater than or equal to zero. As the derived theorem does not involve piecewise expressions, the precision loss caused by the breakpoint issue is avoided. A low-complexity binary-search-based address localization method is proposed to optimize our theorem for hardware implementation. For the optimized scheme, an efficient implemented circuit is also presented. Our scheme’s approximation ability and hardware efficiency are validated through software modeling and FPGA- and ASIC-based experiments. Feedforward neural network-based classification applications demonstrate that building networks with the proposed sigmoid approximator has only a tiny recognition rate loss.

Torque sensors are essential components of robotic joints. In the past, structure optimization of force-sensing elements has been the common approach to improve the performance of the torque sensors. In this work, we demonstrate a torque sensor with bulk metallic glasses as a force-sensing element. Compared with the sensors made of stainless steel and aluminum alloy, the use of bulk metallic glass as a force-sensing element significantly improves sensor sensitivity, linearity, repeatability, hysteresis, and measuring range. Our work not only opens up a new avenue for the application of bulk metallic glasses, but also provides opportunities for enhancing the performance of force/torque sensors through materials optimization.

Graphical abstract:

The dual neural network-based $k$ -winner-take-all (DNN- $k$ WTA) is an analog neural model that is used to identify the $k$ largest numbers from $n$ inputs. Since threshold logic units (TLUs) are key elements in the model, offset voltage drifts in TLUs may affect the operational correctness of a DNN- $k$ WTA network. Previous studies assume that drifts in TLUs follow some particular distributions. This brief considers that only the drift range, given by $[-\Delta, \Delta]$ , is available. We consider two drift cases: time-invariant and time-varying. For the time-invariant case, we show that the state of a DNN- $k$ WTA network converges. The sufficient condition to make a network with the correct operation is given. Furthermore, for uniformly distributed inputs, we prove that the probability that a DNN- $k$ WTA network operates properly is greater than $(1-2\Delta)^{n}$. The aforementioned results are generalized for the time-varying case. In addition, for the time-invariant case, we derive a method to compute the exact convergence time for a given data set. For uniformly distributed inputs, we further derive the mean and variance of the convergence time. The convergence time results give us an idea about the operational speed of the DNN- $k$ WTA model. Finally, simulation experiments have been conducted to validate those theoretical results. [ABSTRACT FROM AUTHOR]

The dual neural network-based $k$ -winner-take-all (DNN- $k$ WTA) is an analog neural model that is used to identify the $k$ largest numbers from $n$ inputs. Since threshold logic units (TLUs) are key elements in the model, offset voltage drifts in TLUs may affect the operational correctness of a DNN- $k$ WTA network. Previous studies assume that drifts in TLUs follow some particular distributions. This brief considers that only the drift range, given by $[-\Delta, \Delta]$ , is available. We consider two drift cases: time-invariant and time-varying. For the time-invariant case, we show that the state of a DNN- $k$ WTA network converges. The sufficient condition to make a network with the correct operation is given. Furthermore, for uniformly distributed inputs, we prove that the probability that a DNN- $k$ WTA network operates properly is greater than $(1-2\Delta)^{n}$ . The aforementioned results are generalized for the time-varying case. In addition, for the time-invariant case, we derive a method to compute the exact convergence time for a given data set. For uniformly distributed inputs, we further derive the mean and variance of the convergence time. The convergence time results give us an idea about the operational speed of the DNN- $k$ WTA model. Finally, simulation experiments have been conducted to validate those theoretical results.

Meniscus suture is an important treatment method for meniscus injury and contributes to the preservation of proprioception, restoration of knee biomechanics and alleviation of progressive osteoarthritis. However, there are few visualized analyses concerning the present studies of meniscus suture. This paper aims to evaluate the global trends, highlights and frontiers of meniscus suture. A bibliometric analysis was conducted based on the results of studies related to meniscus suture from web of science core collection. VOSviewer, GraphPad Prism, Microsoft Excel and R-bibliometrix were utilized for the bibliometric analysis of country and institution distribution, chronological distribution, source journals analysis, prolific authors and institutions analysis, keywords analysis, and reference co-citation analysis. A total of 950 publications on meniscus suture from 177 different sources were retrieved over the set time span. These publications were completed by 3177 authors from 1112 institutions in 54 countries. The United States was the most prolific country with 7960 citations and 348 publications (36.63%). Furumatsu Takayuki acted as the most prolific author (51 publications), while Robert F LaPrade with 1398 citations was the most-cited author. And more papers were published in the core journals, including American Journal of Sports Medicine, Arthroscopy-The Journal of Arthroscopic and Related Surgery, Knee Surgery Sports Traumatology Arthroscopyand Arthroscopy Techniques. Furthermore, “meniscus healing,” “meniscus root tear” seem to be the emerging research hotspots. Notably, the publication trend concerning the all-inside suture technique has been rising during the past decade. The number of research publications on meniscus suture has been continuously risen since 2010. The authors, publications and institutions from the United States and East Asia were still the mainstays in this field. And the all-inside suture may become the mainstream surgical technique in the future, with meniscus healing and meniscus root tears being research highlights recently.

Recently a lot of works have been investigating to find the tenuous groups, i.e., groups with few social interactions and weak relationships among members, for reviewer selection and psycho-educational group formation. However, the metrics (e.g., k-triangle, k-line, and k-tenuity) used to measure the tenuity, require a suitable kvalue to be specified which is difficult for users without background knowledge. Thus, in this paper we formulate the most tenuous group (MTG) query in terms of the group distance and average group distance of a group measuring the tenuity to eliminate the influence of parameter kon the tenuity of the group. To address the MTG problem, we first propose an exact algorithm, namely MTG-VDIS, which takes priority to selecting those vertices whose vertex distance is large, to generate the result group, and also utilizes effective filtering and pruning strategies. Since MTG-VDIS is not fast enough, we design an efficient exact algorithm, called MTG-VDGE, which exploits the degree metric to sort the vertexes and proposes a new combination order, namely degree and reverse based branch and bound (DRBB). MTG-VDGE gives priority to those vertices with small degree. For a large p, we further develop an approximation algorithm, namely MTG-VDLT, which discards candidate attendees with high degree to reduce the number of vertices to be considered. The experimental results on real datasets manifest that the proposed algorithms outperform existing approaches on both efficiency and group tenuity.

An anterior cruciate ligament (ACL) tear is a common musculoskeletal injury with a high incidence. Traditional diagnosis employs magnetic response imaging (MRI), physical testing, or other clinical examination, which relies on complex and expensive medical instruments, or individual doctoral experience. Herein, we propose a wearable displacement sensing system based on a grating-structured triboelectric stretch sensor to diagnose the ACL injuries. The stretch sensor exhibits a high resolution (0.2 mm) and outstanding robustness (over 1,000,000 continuous operation cycles). This system is employed in clinical trial to diagnose ACL injuries. It measures the displacement difference between the affected leg and the healthy leg during Lachman test. And when such a difference is greater than 3 mm, the ACL is considered to be at risk for injury or tear. Compared with the gold standard of arthroscopy, the consistency rate of this wearable diagnostic system reached about 85.7%, which is higher than that of the Kneelax3 arthrometer (78.6%) with a large volume. This shows that the wearable system possesses the feasibility to supplement and improve existing arthrometers for facile diagnosing ACL injuries. It may take a promising step for wearable healthcare.