Hidekatsu Ito

Memory on a chip

The rat hippocampal neurons were cultured on a dish with 64 planer micro electrodes array. Neurons reorganized a functional network, and an external inputs form outer world elicited a reproducible, particular spatiotemporal pattern of evoked action potentials. The patterns were not completely uniform for repeptive stimulation but varied slightly. We previously reported that dissociated rat hippocampal neurons possessed hysterical function in its network dynamics. The spatio-temporal electrical activity of the living neuronal network evoked by paired stimulation is quite different from one evoked by a single stimulation, suggested that the network can “memorize” an existence of precedent stimulation. In this report, the feature of this type of network hysteresis was investigated and the hysteresis was kept for at least 1s and not kept for 10s. In addition, this hysteresis was remarkable only in aged cultures with matured complex network. Network hysteresis depends on internal state of the network, generated by the autonomous activity in self-organized networks.

Fundamental short-term memory of semi-artificial neuronal network

Spatiotemporal pattern of neuronal network activity is a key component of brain information processing. Cultured rat hippocampal neurons on the multielectrodes array dish are suitable for analyzing and manipulating network dynamics and its developmental changes. We applied paired electrical inputs at various inter-stimulus intervals (ISi) and analyzed the spatio-temporal pattern of evoked responses. We found that the pattern of evoked electrical activity was affected by existence of a prior input in the case that ISi of paired stimuli was within 2 s. These results suggest that a semi-artificial neuronal network on a culture dish has a fundamental component of short-term memory, and the origin of this hysteresis is transition among the internal states of the network, undertaken by synaptic transmissions.

Neurorobot Vitroid as a model of brain-body interaction

To mimic biological intelligence, it is critical to elucidate the network dynamics of a neural network. The dissociated culture system possesses a simple network comparing to a whole brain, thus it is suitable for exploration of spatiotemporal dynamics of electrical activity of a neuronal circuit. Cultured neuronal network has no input-output system, so it requires an artificial peripheral system to interact with outer world. We are developing the neurorobot as the model system for biological information processing with vital components and the artificial peripheral system. The behavior of the neuro-robot is determined by the response pattern of neuronal electrical activity evoked by a current stimulation from outer world. In this study, we developed a novel type of neurorobot with Self-Organization Map (SOM) for a neuronal output pattern decoder. The robot with SOM is expected to perform non-stop learning and generation of behavior simultaneously. The spatiotemporal electrical patterns evoked by the inputs according to the value of the IR sensors on the robot body are translated to 64 dimension feature vectors and inputted to the SOM. Then the 64 dimension feature vectors are mapped to a certain winner vector in the 10 × 10 output layer of SOM. Winner nodes are linked to the purposive behaviors adequate to the inputs according to outer phenomenon. Only at the beginning of the behavior, neurorobot SOM selects two winner nodes premisely assigned to the specific inputs for the obstacles near the L and R side of the robot body. We call the process as “seeding”. After the seeding process, the distribution of winner units for the two inputs were separated each other, when the spatiotemporal pattern of electrical activity were not overlapped. In addition, the position of the centers of winner nodes gravities, updated with every input, are almost stable in the output layer of the SOM.

Modification of activity pattern induced by synaptic enhancements in a semi-artificial network of living neurons

Higher brain function such as memory formation was not performed by activity of a single neuron but performed by functions of a complex network of neuronal cells. The simple small-scaled network of neuronal cells is fully suitable for such interactions between neurons. Dissociated neurons form a network depending on their electrical activity and spontaneous activity frequently observed within a week. We cultured a network of dissociated neurons on a culture dish with 64 planer microelectrodes. We induced synaptic enhancement in cultured neuronal networks by exposing to Mg2+-free condition for 20 min. Mg2+-free condition was achieved by exchanging of normal cell external solution to Mg2+-free recording solution. After the induction of synaptic enhancement, we analyzed activity pattern by an autocorrelogram-based and crosscorrelogram-based method. Autocorrelogram of the neuronal activity centralized, suggesting that the accuracy of the periodicity increased. This drastic change was induced within only 20 minutes. Crosscorrelogram shows those network activity changed to be more synchronously than one before exposure to Mg2+-free condition. These results suggest that functional connections in a semi-artificial neuronal network were changed to ones performing enhanced network activity than before. The modification of the spatiotemporal pattern of activity is thought to be a base of memory in vivo experiments. We performed similar phenomenon in this semi-artificial, autonomously reorganized network of neurons. By elucidation of these modified functional connections in neural network, we can find a cue how to control biological memory formation.

Prefrontal Activity Evoked by Transcranial Magnetic Stimulations (TMS) Is Enhanced by Observing the Behavior of Others

Transcranial magnetic stimulation (TMS) is one of non-invasive brain stimulation methods, and is often utilized for rehabilitation combined with repetitive facilitative exercise. The brain activity evoked by TMS was analyzed as functional Near Infrared Spectroscopy (fNIRS) signal from frontal robe. We found that prefrontal blood flow increased by TMS to the motor cortex. In addition, the prefrontal activity was enhanced by observing a video content of an exercise behavior of others, considered to activate mirror neurons. These results suggest the possibility of improving the effect of rehabilitation using TMS by visual priming.

Does Representation of Outer Objects in Living Neuronal Network Synthesize “the concept”?

Cultured living neuronal network with input-output-interface is the useful model to investigate how to autonomously create an internal object. Using living neuronal network cultured on a multielectrodes-array-dish, We elucidated the relationships between activity patterns evoked by inputs to 3 different electrodes. Activity patterns evoked by 3 different inputs were classified into several clusters including patterns evoked by 3 inputs at various ratios. Repeated input in stable temporal pattern led the clusters to fuse into single cluster. These results indicated that the Cultured living neuronal network (CLNN) separates and classifies the input from outer world and distinguished clusters are autonomously formed.

Relationship between evoked electrical responses and robotic behavior analyzed by Self-Organization Map

Toward neuroprosthetic technology, it is critical that a simple model system for interaction between brain and electric devices. For this purpose, we developed neurorobot system, Vitroid, equipped with a living neuronal network and a miniature moving robot as a body of the neurorobot. Self-Organization-Map (SOM) was employed as a generator for behavior of Vitroid. SOM was designed to map a high-dimensional feature vector to a 2-dimentional vector as the winner unit in output layer of SOM. Furthermore, neighboring units were assigned to resemble input vectors. Thus, SOM also performs pattern classifying analysis for inputted feature vector of neuronal activity. Cultured neuronal networks on Multi-Electrodes-Array (MEA) dish was alternately stimulated by two different electrodes. SOM mapped patterns induced by electrical stimulation to a 30 × 30 - 2D output layer. Only in the first step of the learning, SOM is forced to select a specific winner unit previously assigned in order to associate specific behaviors. We call this process “Seeding”. After seeding process, the winner-units correspond to the response patterns induced by two different stimuli were separately mapped. We confirmed that response patterns by two different electrical stimuli could be classified and they were almost stable. Furthermore, it revealed that spontaneous activity and evoked response shared the same patterns, suggesting that the internal autonomous activity is not only a noise, but is almost equivalent to a meaningful response. We also succeeded in collision avoidance of Vitroid by SOM-based behavior generator.

Relationship between inter-stimulus-intervals and intervals of autonomous activities in a neuronal network

To investigate relationships between neuronal network activity and electrical stimulus, we analyzed autonomous activity before and after electrical stimulus. Recordings of autonomous activity were performed using dissociated culture of rat hippocampal neurons on a multi-electrodes array (MEA) dish. Single stimulus and pared stimuli were applied to a cultured neuronal network. Single stimulus was applied every 1 min, and paired stimuli was performed by two sequential stimuli every 1 min. As a result, the patterns of synchronized activities of a neuronal network were changed after stimulus. Especially, long range synchronous activities were induced by paired stimuli. When 1 s inter-stimulus-intervals (ISI) and 1.5 s ISI paired stimuli are applied to a neuronal network, relatively long range synchronous activities expressed in case of 1.5 s ISI. Temporal synchronous activity of neuronal network is changed according to inter-stimulus-intervals (ISI) of electrical stimulus. In other words, dissociated neuronal network can maintain given information in temporal pattern and a certain type of an information maintenance mechanism was considered to be implemented in a semi-artificial dissociated neuronal network. The result is useful toward manipulation technology of neuronal activity in a brain system.

The effects of transient abolishment of electrical activity on dynamics in a dissociated neuronal network

The higher-order functions of brain depend on neuronal network electrical activity with complex spatiotemporal patterns. The background activity, such as spontaneous electrical activity observed in a dissociated culture of rat hippocampal neurons, is considered to be a fundamental component of internal state of the living neuronal network. The spontaneous activity has dynamic and complicated spatiotemporal activity patterns, thus it is uncertain that transiently discontinued activity pattern recovers to the same activity state. We elucidated the stability of the internal state of network activity after a transient abolishment of neuronal electrical activity. As a result, spontaneous activity pattern became to be bursting and intermittent pattern after the transient block of electrical activity. In addition, spontaneous neuronal activity increased in frequency. These modified patterns lasted for hours then gradually returned to the initial state. These results suggest that steady state of spontaneous activity is a result of complex equilibrium of interactions between neurons. Thus, it is required to consider the influence of the inhibition itself when the neural activity is stopped by pharmacological manipulations.

Stability of neuronal electrical activity pattern evoked by two inputs stimulation

Rat hippocampal dissociated culture on multi electrodes array dish (MED) is useful as minimalized brain model to investigation of principles of brain information processing. Rat hippocampal neuronal cells were cultured on MED with 64 microelectrodes and they reconstructed a complex network. We analyzed stability of neuronal electrical activity pattern after two distinct electrical stimuli. Distances between averaged spike pattern and spike pattern after electrical stimulation for summarizing the activity pattern to temporal axis direction and trial (sweep) axis direction. The distances of the analysis for summarized to temporal axis direction increased immediately after stimulation and after it, recovered to standard level. Moreover, stabilizing points emerge every 100 ms on both analysis for summarizing the activity pattern to temporal axis direction and trial axis direction. These results suggest that the neural network buttress by electrical stimulation and it be complex by two-point inputs.