Broadband sensory networks with locally stored responsivities for neuromorphic machine vision.

As the most promising candidates for the implementation of in-sensor computing, retinomorphic vision sensors can constitute built-in neural networks and directly implement multiply-and-accumulation operations using responsivities as the weights. However, existing retinomorphic vision sensors mainly use a sustained gate bias to maintain the responsivity due to its volatile nature. Here, we propose an ion-induced localized-field strategy to develop retinomorphic vision sensors with nonvolatile tunable responsivity in both positive and negative regimes and construct a broadband and reconfigurable sensory network with locally stored weights to implement in-sensor convolutional processing in spectral range of 400 to 1800 nanometers. In addition to in-sensor computing, this retinomorphic device can implement in-memory computing benefiting from the nonvolatile tunable conductance, and a complete neuromorphic visual system involving front-end in-sensor computing and back-end in-memory computing architectures has been constructed, executing supervised and unsupervised learning tasks as demonstrations. This work paves the way for the development of high-speed and low-power neuromorphic machine vision for time-critical and data-intensive applications.

Holistically-Attracted Wireframe Parsing: From Supervised to Self-Supervised Learning.

This article presents Holistically-Attracted Wireframe Parsing (HAWP), a method for geometric analysis of 2D images containing wireframes formed by line segments and junctions. HAWP utilizes a parsimonious Holistic Attraction (HAT) field representation that encodes line segments using a closed-form 4D geometric vector field. The proposed HAWP consists of three sequential components empowered by end-to-end and HAT-driven designs: 1) generating a dense set of line segments from HAT fields and endpoint proposals from heatmaps, 2) binding the dense line segments to sparse endpoint proposals to produce initial wireframes, and 3) filtering false positive proposals through a novel endpoint-decoupled line-of-interest aligning (EPD LOIAlign) module that captures the co-occurrence between endpoint proposals and HAT fields for better verification. Thanks to our novel designs, HAWPv2 shows strong performance in fully supervised learning, while HAWPv3 excels in self-supervised learning, achieving superior repeatability scores and efficient training (24 GPU hours on a single GPU). Furthermore, HAWPv3 exhibits a promising potential for wireframe parsing in out-of-distribution images without providing ground truth labels of wireframes.

Low-voltage ultrafast nonvolatile memory via direct charge injection through a threshold resistive-switching layer.

The explosion in demand for massive data processing and storage requires revolutionary memory technologies featuring ultrahigh speed, ultralong retention, ultrahigh capacity and ultralow energy consumption. Although a breakthrough in ultrafast floating-gate memory has been achieved very recently, it still suffers a high operation voltage (tens of volts) due to the Fowler-Nordheim tunnelling mechanism. It is still a great challenge to realize ultrafast nonvolatile storage with low operation voltage. Here we propose a floating-gate memory with a structure of MoS2/hBN/MoS2/graphdiyne oxide/WSe2, in which a threshold switching layer, graphdiyne oxide, instead of a dielectric blocking layer in conventional floating-gate memories, is used to connect the floating gate and control gate. The volatile threshold switching characteristic of graphdiyne oxide allows the direct charge injection from control gate to floating gate by applying a nanosecond voltage pulse (20 ns) with low magnitude (2 V), and restricts the injected charges in floating gate for a long-term retention (10 years) after the pulse. The high operation speed and low voltage endow the device with an ultralow energy consumption of 10 fJ. These results demonstrate a new strategy to develop next-generation high-speed low-energy nonvolatile memory.

Presynaptic NMDARs on spinal nociceptor terminals state-dependently modulate synaptic transmission and pain.

Postsynaptic NMDARs at spinal synapses are required for postsynaptic long-term potentiation and chronic pain. However, how presynaptic NMDARs (PreNMDARs) in spinal nociceptor terminals control presynaptic plasticity and pain hypersensitivity has remained unclear. Here we report that PreNMDARs in spinal nociceptor terminals modulate synaptic transmission in a nociceptive tone-dependent manner. PreNMDARs depresses presynaptic transmission in basal state, while paradoxically causing presynaptic potentiation upon injury. This state-dependent modulation is dependent on Ca2+ influx via PreNMDARs. Small conductance Ca2+-activated K+ (SK) channels are responsible for PreNMDARs-mediated synaptic depression. Rather, tissue inflammation induces PreNMDARs-PKG-I-dependent BDNF secretion from spinal nociceptor terminals, leading to SK channels downregulation, which in turn converts presynaptic depression to potentiation. Our findings shed light on the state-dependent characteristics of PreNMDARs in spinal nociceptor terminals on modulating nociceptive transmission and revealed a mechanism underlying state-dependent transition. Moreover, we identify PreNMDARs in spinal nociceptor terminals as key constituents of activity-dependent pain sensitization.

Extracellular matrix protein laminin ß1 regulates pain sensitivity and anxiodepression-like behaviors in mice.

Patients with neuropathic pain often experience comorbid psychiatric disorders. Cellular plasticity in the anterior cingulate cortex (ACC) is assumed to be a critical interface for pain perception and emotion. However, substantial efforts have thus far been focused on the intracellular mechanisms of plasticity rather than the extracellular alterations that might trigger and facilitate intracellular changes. Laminin, a key element of the extracellular matrix (ECM), consists of one α-, one ß-, and one γ-chain and is implicated in several pathophysiological processes. Here, we showed in mice that laminin ß1 (LAMB1) in the ACC was significantly downregulated upon peripheral neuropathy. Knockdown of LAMB1 in the ACC exacerbated pain sensitivity and induced anxiety and depression. Mechanistic analysis revealed that loss of LAMB1 caused actin dysregulation via interaction with integrin ß1 and the subsequent Src-dependent RhoA/LIMK/cofilin pathway, leading to increased presynaptic transmitter release probability and abnormal postsynaptic spine remodeling, which in turn orchestrated the structural and functional plasticity of pyramidal neurons and eventually resulted in pain hypersensitivity and anxiodepression. This study sheds new light on the functional capability of ECM LAMB1 in modulating pain plasticity and identifies a mechanism that conveys extracellular alterations to intracellular plasticity. Moreover, we identified cingulate LAMB1/integrin ß1 signaling as a promising therapeutic target for the treatment of neuropathic pain and associated anxiodepression.

Learning Regional Attraction for Line Segment Detection.

This paper presents regional attraction of line segment maps, and hereby poses the problem of line segment detection (LSD) as a problem of region coloring. Given a line segment map, the proposed regional attraction first establishes the relationship between line segments and regions in the image lattice. Based on this, the line segment map is equivalently transformed to an attraction field map (AFM), which can be remapped to a set of line segments without loss of information. Accordingly, we develop an end-to-end framework to learn attraction field maps for raw input images, followed by a squeeze module to detect line segments. Apart from existing works, the proposed detector properly handles the local ambiguity and does not rely on the accurate identification of edge pixels. Comprehensive experiments on the Wireframe dataset and the YorkUrban dataset demonstrate the superiority of our method. In particular, we achieve an F-measure of 0.831 on the Wireframe dataset, advancing the state-of-the-art performance by 10.3 percent.

Nociceptor-localized cGMP-dependent protein kinase I is a critical generator for central sensitization and neuropathic pain.

Patients with neuropathic pain often experience exaggerated pain and anxiety. Central sensitization has been linked with the maintenance of neuropathic pain and may become an autonomous pain generator. Conversely, emerging evidence accumulated that central sensitization is initiated and maintained by ongoing nociceptive primary afferent inputs. However, it remains elusive what mechanisms underlie this phenomenon and which peripheral candidate contributes to central sensitization that accounts for pain hypersensitivity and pain-related anxiety. Previous studies have implicated peripherally localized cGMP-dependent protein kinase I (PKG-I) in plasticity of nociceptors and spinal synaptic transmission as well as inflammatory hyperalgesia. However, whether peripheral PKG-I contributes to cortical plasticity and hence maintains nerve injury-induced pain hypersensitivity and anxiety is unknown. Here, we demonstrated significant upregulation of PKG-I in ipsilateral L3 dorsal root ganglia (DRG), no change in L4 DRG, and downregulation in L5 DRG upon spared nerve injury. Genetic ablation of PKG-I specifically in nociceptors or post-treatment with intervertebral foramen injection of PKG-I antagonist, KT5823, attenuated the development and maintenance of spared nerve injury-induced bilateral pain hypersensitivity and anxiety. Mechanistic analysis revealed that activation of PKG-I in nociceptors is responsible for synaptic potentiation in the anterior cingulate cortex upon peripheral neuropathy through presynaptic mechanisms involving brain-derived neurotropic factor signaling. Our results revealed that PKG-I expressed in nociceptors is a key determinant for cingulate synaptic plasticity after nerve injury, which contributes to the maintenance of pain hypersensitivity and anxiety. Thereby, this study presents a strong basis for opening up a novel therapeutic target, PKG-I, in nociceptors for treatment of comorbidity of neuropathic pain and anxiety with least side effects.

Tweety-Homolog 1 Facilitates Pain via Enhancement of Nociceptor Excitability and Spinal Synaptic Transmission.

Tweety-homolog 1 (Ttyh1) is expressed in neural tissue and has been implicated in the generation of several brain diseases. However, its functional significance in pain processing is not understood. By disrupting the gene encoding Ttyh1, we found a loss of Ttyh1 in nociceptors and their central terminals in Ttyh1-deficient mice, along with a reduction in nociceptor excitability and synaptic transmission at identified synapses between nociceptors and spinal neurons projecting to the periaqueductal grey (PAG) in the basal state. More importantly, the peripheral inflammation-evoked nociceptor hyperexcitability and spinal synaptic potentiation recorded in spinal-PAG projection neurons were compromised in Ttyh1-deficient mice. Analysis of the paired-pulse ratio and miniature excitatory postsynaptic currents indicated a role of presynaptic Ttyh1 from spinal nociceptor terminals in the regulation of neurotransmitter release. Interfering with Ttyh1 specifically in nociceptors produces a comparable pain relief. Thus, in this study we demonstrated that Ttyh1 is a critical determinant of acute nociception and pain sensitization caused by peripheral inflammation.

[An investigation of severe neonatal hyperbilirubinemia in 13 hospitals of Jiangsu Province, China].

OBJECTIVE: To investigate the incidence of severe neonatal hyperbilirubinemia and the management on the treatment and follow-up of this disease in Jiangsu Province, China. METHODS: The neonates with severe hyperbilirubinemia who were admitted to 13 hospitals in Jiangsu Province from January to December, 2018, were enrolled as subjects. A retrospective analysis was performed on their mediacal data and follow-up data. RESULTS: In 2018, 740 neonates with severe hyperbilirubinemia were reported from the 13 hospitals in Jiangsu Province, accounting for 2.70% (740/27 386) of the total number of neonates admitted to the department of neonatology. Among these neonates, 620 (83.8%) had severe hyperbilirubinemia, 106 (14.3%) had extremely severe hyperbilirubinemia, and 14 (1.9%) had hazardous hyperbilirubinemia. Four neonates (0.5%) were diagnosed with acute bilirubin encephalopathy. A total of 484 neonates (65.4%) were readmitted due to severe hyperbilirubinemia after discharge from the delivery institution, with a median age of 7 days, among whom 214 (44.2%) were followed up for jaundice at the outpatient service before readmission, with a median age of 6 days at the first time of outpatient examination. During hospitalization, 211 neonates (28.5%) underwent cranial MRI examinations, among whom 85 (40.3%) had high T1WI signal in the bilateral basal ganglia and the globus pallidus; 238 neonates (32.2%) underwent brainstem auditory evoked potential examinations, among whom 14 (5.9%) passed only at one side and 7 (2.9%) failed at both sides. The 17 neonates with acute bilirubin encephalopathy or hazardous hyperbilirubinemia were followed up. Except one neonate was lost to follow-up, and there were no abnormal neurological symptoms in the other neonates. CONCLUSIONS: Neonates with severe hyperbilirubinemia account for a relatively high proportion of the total number of neonates in the department of neonatology. Jaundice monitoring and management after discharge from delivery institutions need to be strengthened. For neonates with severe hyperbilirubinemia, relevant examinations should be carried out more comprehensively during hospitalization and these neonates should be followed up comprehensively and systematically after discharge.

TRPC1/4/5 channels contribute to morphine-induced analgesic tolerance and hyperalgesia by enhancing spinal synaptic potentiation and structural plasticity.

Opioid analgesics remain the mainstay for managing intractable chronic pain, but their use is limited by detrimental side effects such as analgesic tolerance and hyperalgesia. Calcium-dependent synaptic plasticity is a key determinant in opiates tolerance and hyperalgesia. However, the exact substrates for this calcium-dependent synaptic plasticity in mediating these maladaptive processes are largely unknown. Canonical transient receptor potential 1, 4, and 5 (TRPC1, 4, 5) proteins assemble into heteromultimeric nonselective cation channels with high Ca2+ permeability and influence various neuronal functions. However, whether and how TRPC1/4/5 channels contribute to the development of opiates tolerance and hyperalgesia remains elusive. Here, we show that TRPC1/4/5 channels contribute to the generation of morphine tolerance and hyperalgesia. Chronic morphine exposure leads to upregulation of TRPC1/4/5 channels in the spinal cord. Spinally expressed TRPC1, TPRC4, and TRPC5 are required for chronic morphine-induced synaptic long-term potentiation (LTP) as well as remodeling of synaptic spines in the dorsal horn, thereby orchestrating functional and structural plasticity during the course of morphine-induced hyperalgesia and tolerance. These effects are attributed to TRPC1/4/5-mediated Ca2+ elevation in the spinal dorsal horn induced by chronic morphine treatment. This study identifies TRPC1/4/5 channels as a promising novel target to prevent the unwanted morphine tolerance and hyperalgesia.

A Functional Representation for Graph Matching.

Graph matching is an important and persistent problem in computer vision and pattern recognition for finding node-to-node correspondence between graphs. However, graph matching that incorporates pairwise constraints can be formulated as a quadratic assignment problem (QAP), which is NP-complete and results in intrinsic computational difficulties. This paper presents a functional representation for graph matching (FRGM) that aims to provide more geometric insights on the problem and reduce the space and time complexities. To achieve these goals, we represent each graph by a linear function space equipped with a functional such as inner product or metric, that has an explicit geometric meaning. Consequently, the correspondence matrix between graphs can be represented as a linear representation map. Furthermore, this map can be reformulated as a new parameterization for matching graphs in Euclidean space such that it is consistent with graphs under rigid or nonrigid deformations. This allows us to estimate the correspondence matrix and geometric deformations simultaneously. We use the representation of edge-attributes rather than the affinity matrix to reduce the space complexity and propose an efficient optimization strategy to reduce the time complexity. The experimental results on both synthetic and real-world datasets show that the FRGM can achieve state-of-the-art performance.

[Studies on triterpenoid constituents from Rabdosia japonica var. galaucocalyx].

OBJECTIVE: To study the chemical constitutes in Rabdosia japonica var. galaucocalyx . METHOD: The compounds were isolated by nomal phase silica gel chromatography. The structures were identified by physical and spectral data. RESULT: Nine triterpenoids were isolated and identified as friedelin (I), 3beta, 28-dihydroxy-ursane (II), ursolic acid (III), 3beta-acetyloxy-ursolic acid (IV), 2alpha, 3alpha-dihydroxy-urs-12-en-28-oic acid (V), 2alpha, 3alpha, 23-trihydroxy-urs-12-en-28-oic acid (VI), oleanic acid (VII), beta-sitosterol (VIII), daucousterol (IX). CONCLUSION: Compounds I, II, IV, V and VI were isolated from R. japonica var. galaucocalyx for the first time.