Reservoir computing systems utilize dynamic reservoirs having short-term memory to project features from the temporal inputs into a high-dimensional feature space. A readout function layer can then effectively analyze the projected features for tasks, such as classification and time-series analysis. The system can efficiently compute complex and temporal data with low-training cost, since only the readout function needs to be trained. Here we experimentally implement a reservoir computing system using a dynamic memristor array. We show that the internal ionic dynamic processes of memristors allow the memristor-based reservoir to directly process information in the temporal domain, and demonstrate that even a small hardware system with only 88 memristors can already be used for tasks, such as handwritten digit recognition. The system is also used to experimentally solve a second-order nonlinear task, and can successfully predict the expected output without knowing the form of the original dynamic transfer function.

Memristors have been extensively studied for data storage and low-power computation applications. In this study, we show that memristors offer more than simple resistance change. Specifically, the dynamic evolutions of internal state variables allow an oxide-based memristor to exhibit Ca(2+)-like dynamics that natively encode timing information and regulate synaptic weights. Such a device can be modeled as a second-order memristor and allow the implementation of critical synaptic functions realistically using simple spike forms based solely on spike activity.

Sparse representation of information provides a powerful means to perform feature extraction on high-dimensional data and is of broad interest for applications in signal processing, computer vision, object recognition and neurobiology. Sparse coding is also believed to be a key mechanism by which biological neural systems can efficiently process a large amount of complex sensory data while consuming very little power. Here, we report the experimental implementation of sparse coding algorithms in a bio-inspired approach using a 32 × 32 crossbar array of analog memristors. This network enables efficient implementation of pattern matching and lateral neuron inhibition and allows input data to be sparsely encoded using neuron activities and stored dictionary elements. Different dictionary sets can be trained and stored in the same system, depending on the nature of the input signals. Using the sparse coding algorithm, we also perform natural image processing based on a learned dictionary.

wileyonlinelibrary.comspike-timing-dependent plasticity (STDP) using various types of memristors. [ 12,[16][17][18][19][20][21] However, in these studies, the synaptic learning rules were implemented phenomenologically by engineering the duration or amplitude of the overlapping programming pulses from the pre-and postsynaptic neurons. [ 12,[18][19][20][21][22] The phenomenological nature of this approach means that different programming pulses have to be manually designed to implement the desired synaptic behaviors. However, in biology, the apparently different learning rules have been shown to be specifi c effects driven by internal molecular dynamic processes under stimulation. [23][24][25] Consequently, manually designing a system to specifically target only certain effects, but not their cause, can easily miss other important aspects that make the system functional.In a previous study, we showed that by employing multiple internal state variables (e.g., temperature and conduction fi lament size), a second-order memristor can be obtained which allows biorealistic implementation of several synaptic learning rules-notably spike-timing-dependent plasticity. [ 26 ] Here, we show that a second-order memristor can also be implemented by utilizing the different time scales of internal ionic dynamics in oxide-based memristors, leading to the natural implementation of several types of important synaptic behaviors. We show that an oxide-based memristor may be described by two state variables-one ( w c ) directly determines the device conductance (weight) and the other ( w m ) affects the dynamics of the fi rst (conductance) state variable. Specifi cally in our device system, w c represents the area of the conducting channel region in the oxide memristor thus directly affecting the device conductance, while w m represents the oxygen vacancy mobility in the fi lm which directly affects the dynamics of w c but only indirectly modulates the device conductance. Within this secondorder memristor framework, the device long-term state can be shown to be controlled by activities at much shorter time scales. Specifi cally, the natural decay of the state variable w m provides an internal timing and modulation mechanism analogous to that exhibited by Ca 2+ concentration, [23][24][25] and enables the memristor to exhibit important rate-and timing-dependent behaviors at both short-term such as pair-pulse facilitation (PPF) [ 27 ] and long-term such as STDP [ 28 ] using simple, nonoverlapping spike signals. The experimental observations can in turn be quantitatively explained using a simple dynamic device model including the two state variables, and facilitates large-scale simulation and implementation of memristor-based neuromorphic systems.Memristors have attracted broad interest as a promising candidate for future memory and computing applications. Particularly, it is believed that memristors can effectively implement synaptic functions and enable effi cient neuromorphic systems. Most previous studies, however, focus on implementin...

No abstract

No abstract

Presbycusis (PC) is characterized by bilateral sensorineural hearing loss at high frequencies and speech-perception difficulties in noisy environments and has a strikingly detrimental impact on cognitive function. As the neural consequences of PC may involve the whole brain, we hypothesized that patients with PC would show structural alterations not only in the auditory cortex but also in the cortexes involved in cognitive function. The purpose of this study was to use surface-based morphometry (SBM) analysis to elucidate whole-brain structural differences between patients with PC and age-matched normal hearing controls. Three-dimensional T1-weighted MR images of 26 patients with mild PC and 26 age-, sex- and education-matched healthy controls (HCs) were acquired. All participants underwent a battery of neuropsychological tests. Our results revealed gray matter atrophy in several auditory cortical areas, nodes of the default mode network (DMN), including the bilateral precuneus and inferior parietal lobule, the right posterior cingulate cortex (PCC), and the right insula of patients with PC compared to that in the HCs. Our findings also revealed that hearing loss was associated with reduced gray matter volume in the right primary auditory cortex of patients with PC. Moreover, structural alterations in the nodes of the DMN were associated with cognitive impairments in PC patients. Additionally, this study provides evidence that a thicker right insula is associated with better speech perception in patients with PC. Based on these findings, we argue that the onset of PC seems to trigger its own cascade of conditions, including a need for increased cognitive resources during speech comprehension, which might lead to auditory and cognition-related cortical reorganization.

Gamma-aminobutyric acid (GABA) is the main inhibitory neurotransmitter in the central auditory system. Altered GABAergic neurotransmission has been found in both the inferior colliculus and the auditory cortex in animal models of presbycusis. Edited magnetic resonance spectroscopy (MRS), using the MEGA-PRESS sequence, is the most widely used technique for detecting GABA in the human brain. However, to date there has been a paucity of studies exploring changes to the GABA concentrations in the auditory region of patients with presbycusis. In this study, sixteen patients with presbycusis (5 males/11 females, mean age 63.1 ± 2.6 years) and twenty healthy controls (6 males/14 females, mean age 62.5 ± 2.3 years) underwent audiological and MRS examinations. Pure tone audiometry from 0.125 to 8 KHz and tympanometry were used to assess the hearing abilities of all subjects. The pure tone average (PTA; the average of hearing thresholds at 0.5, 1, 2, and 4 kHz) was calculated. The MEGA-PRESS sequence was used to measure GABA+ concentrations in 4 × 3 × 3 cm3 volumes centered on the left and right Heschl’s gyri. GABA+ concentrations were significantly lower in the presbycusis group compared to the control group (left auditory regions: p = 0.002, right auditory regions: p = 0.008). Significant negative correlations were observed between PTA and GABA+ concentrations in the presbycusis group (r = −0.57, p = 0.02), while a similar trend was found in the control group (r = −0.40, p = 0.08). These results are consistent with a hypothesis of dysfunctional GABAergic neurotransmission in the central auditory system in presbycusis, and suggest a potential treatment target for presbycusis.