東北大学研究者紹介

研究者詳細

ホーム

日本語 English

研究者略歴

基本情報
経歴
学歴

研究活動

論文
MISC

教育活動

貢献活動

その他

顔写真

スミ　タクマ

住　拓磨

Takuma Sumi

所属

高等研究機構材料科学高等研究所　数学連携グループ

職名

助教

学位

博士（医工学）（東北大学）
修士（医工学）（東北大学）

researchmap

https://researchmap.jp/takuma_sumi

J-GLOBAL ID

202301003080673334

e-Rad 研究者番号

31002545

ORCID

https://orcid.org/0000-0001-9506-5226

経歴 2

2023年10月～継続中

東北大学　材料科学高等研究所　助教
2021年4月～ 2023年9月

日本学術振興会特別研究員(DC1)

学歴 3

東北大学　大学院医工学研究科

2020年10月～ 2023年9月
東北大学　大学院医工学研究科

2019年4月～ 2020年9月
東北大学　工学部　電気情報物理工学科

2015年4月～ 2019年3月

論文 8

Online supervised learning of temporal patterns in biological neural networks under feedback control

Yuki Sono, Hideaki Yamamoto, Yusei Nishi, Takuma Sumi, Yuya Sato, Ayumi Hirano-Iwata, Yuichi Katori, Shigeo Sato

2025年8月12日

DOI： 10.1101/2025.08.10.669034 　
Modular architecture confers robustness to damage and facilitates recovery in spiking neural networks modeling in vitro neurons 査読有り

Takuma Sumi, Akke Mats Houben, Hideaki Yamamoto, Hideyuki Kato, Yuichi Katori, Jordi Soriano, Ayumi Hirano-Iwata

Frontiers in Neuroscience　19　1570783　2025年6月19日
出版者・発行元： Frontiers Media SA
DOI： 10.3389/fnins.2025.1570783 　

ISSN：1662-453X

eISSN：1662-453X

詳細を見る詳細を閉じる

Impaired brain function is restored following injury through dynamic processes that involve synaptic plasticity. This restoration is supported by the brain’s inherent modular organization, which promotes functional separation and redundancy. However, it remains unclear how modular structure interacts with synaptic plasticity to define damage response and recovery efficiency. In this work, we numerically modeled the response and recovery to damage of a neuronal network in vitro bearing a modular structure. The simulations aimed at capturing experimental observations in cultured neurons with modular traits which were physically disconnected through a focal lesion. The damage reduced the frequency of spontaneous collective activity events in the cultures, which recovered to pre-damage levels within 24 h. We rationalized this recovery in the numerical simulations by considering a plasticity mechanism based on spike-timing-dependent plasticity, a form of synaptic plasticity that modifies synaptic strength based on the relative timing of pre- and postsynaptic spikes. We observed that the in silico numerical model effectively captured the decline and subsequent recovery of spontaneous activity following the injury. The model supports that the combination of modularity and plasticity confers robustness to the damaged neuronal network by preventing the total loss of spontaneous network-wide activity and facilitating recovery. Additionally, by using our model within the reservoir computing framework, we show that information representation in the neuronal network improves with the recovery of network-wide activity.
In silico modeling of reservoir-based predictive coding in biological neuronal networks on microelectrode arrays 査読有り

Yuya Sato, Hideaki Yamamoto, Yoshitaka Ishikawa, Takuma Sumi, Yuki Sono, Shigeo Sato, Yuichi Katori, Ayumi Hirano-Iwata

Japanese Journal of Applied Physics　63　(10)　108001　2024年10月1日
出版者・発行元： IOP Publishing
DOI： 10.35848/1347-4065/ad7ec1 　

ISSN：0021-4922

eISSN：1347-4065

詳細を見る詳細を閉じる

Abstract Reservoir computing and predictive coding together yield a computational model for exploring how neuronal dynamics in the mammalian cortex underpin temporal signal processing. Here, we construct an in-silico model of biological neuronal networks grown on microelectrode arrays and explore their computing capabilities through a sine wave prediction task in a reservoir-based predictive coding framework. Our results show that the time interval between stimulation pulses is a critical determinant of task performance. Additionally, under a fixed feedback latency, pulse amplitude modulation is a favorable encoding scheme for input signals. These findings provide practical guidelines for future implementation of the model in biological experiments.
Integrating predictive coding with reservoir computing in spiking neural network model of cultured neurons 査読有り

Yoshitaka Ishikawa, Takumi Shinkawa, Takuma Sumi, Hideyuki Kato, Hideaki Yamamoto, Yuichi Katori

Nonlinear Theory and Its Applications, IEICE　15　(2)　432-442　2024年
出版者・発行元： Institute of Electronics, Information and Communications Engineers (IEICE)
DOI： 10.1587/nolta.15.432 　

eISSN：2185-4106
Biological neurons act as generalization filters in reservoir computing. 国際誌査読有り

Takuma Sumi, Hideaki Yamamoto, Yuichi Katori, Koki Ito, Satoshi Moriya, Tomohiro Konno, Shigeo Sato, Ayumi Hirano-Iwata

Proceedings of the National Academy of Sciences of the United States of America　120　(25)　e2217008120　2023年6月20日

DOI： 10.1073/pnas.2217008120 　

詳細を見る詳細を閉じる

Reservoir computing is a machine learning paradigm that transforms the transient dynamics of high-dimensional nonlinear systems for processing time-series data. Although the paradigm was initially proposed to model information processing in the mammalian cortex, it remains unclear how the nonrandom network architecture, such as the modular architecture, in the cortex integrates with the biophysics of living neurons to characterize the function of biological neuronal networks (BNNs). Here, we used optogenetics and calcium imaging to record the multicellular responses of cultured BNNs and employed the reservoir computing framework to decode their computational capabilities. Micropatterned substrates were used to embed the modular architecture in the BNNs. We first show that the dynamics of modular BNNs in response to static inputs can be classified with a linear decoder and that the modularity of the BNNs positively correlates with the classification accuracy. We then used a timer task to verify that BNNs possess a short-term memory of several 100 ms and finally show that this property can be exploited for spoken digit classification. Interestingly, BNN-based reservoirs allow categorical learning, wherein a network trained on one dataset can be used to classify separate datasets of the same category. Such classification was not possible when the inputs were directly decoded by a linear decoder, suggesting that BNNs act as a generalization filter to improve reservoir computing performance. Our findings pave the way toward a mechanistic understanding of information representation within BNNs and build future expectations toward the realization of physical reservoir computing systems based on BNNs.
超軟ゲル表面を用いた神経細胞の生体模倣培養技術の開発査読有り

住拓磨, 山本英明, 平野愛弓

表面と真空　63　(6)　298-303　2020年6月10日
出版者・発行元：公益社団法人日本表面真空学会
DOI： 10.1380/vss.63.298 　

ISSN：2433-5835

eISSN：2433-5843
Suppression of hypersynchronous network activity in cultured cortical neurons using an ultrasoft silicone scaffold. 国際誌査読有り

Takuma Sumi, Hideaki Yamamoto, Ayumi Hirano-Iwata

Soft matter　16　(13)　3195-3202　2020年4月1日

DOI： 10.1039/c9sm02432h 　

詳細を見る詳細を閉じる

The spontaneous activity pattern of cortical neurons in dissociated culture is characterized by burst firing that is highly synchronized among a wide population of cells. The degree of synchrony, however, is excessively higher than that in cortical tissues. Here, we employed polydimethylsiloxane (PDMS) elastomers to establish a novel system for culturing neurons on a scaffold with an elastic modulus resembling brain tissue, and investigated the effect of the scaffold's elasticity on network activity patterns in cultured rat cortical neurons. Using whole-cell patch clamp to assess the scaffold effect on the development of synaptic connections, we found that the amplitude of excitatory postsynaptic current, as well as the frequency of spontaneous transmissions, was reduced in neuronal networks grown on an ultrasoft PDMS with an elastic modulus of 0.5 kPa. Furthermore, the ultrasoft scaffold was found to suppress neural correlations in the spontaneous activity of the cultured neuronal network. The dose of GsMTx-4, an antagonist of stretch-activated cation channels (SACs), required to reduce the generation of the events below 1.0 event per min on PDMS substrates was lower than that for neurons on a glass substrate. This suggests that the difference in the baseline level of SAC activation is a molecular mechanism underlying the alteration in neuronal network activity depending on scaffold stiffness. Our results demonstrate the potential application of PDMS with biomimetic elasticity as cell-culture scaffold for bridging the in vivo-in vitro gap in neuronal systems.
Ultrasoft Silicone Gel as a Biomimetic Passivation Layer in Inkjet-Printed 3D MEA Devices. 国際誌査読有り

Hideaki Yamamoto, Leroy Grob, Takuma Sumi, Kazuhiro Oiwa, Ayumi Hirano-Iwata, Bernhard Wolfrum

Advanced biosystems　3　(9)　e1900130　2019年9月

DOI： 10.1002/adbi.201900130 　

詳細を見る詳細を閉じる

Multielectrode arrays (MEAs) are versatile tools that are used for chronic recording and stimulation of neural cells and tissues. Driven by the recent progress in understanding of how neuronal growth and function respond to scaffold stiffness, development of MEAs with a soft cell-to-device interface has gained importance not only for in vivo but also for in vitro applications. However, the passivation layer, which constitutes the majority of the cell-device interface, is typically prepared with stiff materials. Herein, a fabrication of an MEA device with an ultrasoft passivation layer is described, which takes advantage of inkjet printing and a polydimethylsiloxane (PDMS) gel with a stiffness comparable to that of the brain. The major challenge in using the PDMS gel is that it cannot be patterned to expose the sensing area of the electrode. This issue is resolved by printing 3D micropillars at the electrode tip. Primary cortical neurons are grown on the fabricated device, and effective stimulation of the culture confirms functional cell-device coupling. The 3D MEA device with an ultrasoft interface provides a novel platform for investigating evoked activity and drug responses of living neuronal networks cultured in a biomimetic environment for both fundamental research and pharmaceutical applications.

︎全件表示 ︎最初の5件までを表示

MISC 25

Reservoir computing with the Kuramoto model

Hayato Chiba, Koichi Taniguchi, Takuma Sumi

2024年7月23日

詳細を見る詳細を閉じる

Reservoir computing aims to achieve high-performance and low-cost machine learning with a dynamical system as a reservoir. However, in general, there are almost no theoretical guidelines for its high-performance or optimality. This paper focuses on the reservoir computing with the Kuramoto model and theoretically reveals its approximation ability. The main result provides an explicit expression of the dynamics of the Kuramoto reservoir by using the order parameters. Thus, the output of the reservoir computing is expressed as a linear combination of the order parameters. As a corollary, sufficient conditions on hyperparameters are obtained so that the set of the order parameters gives the complete basis of the Lebesgue space. This implies that the Kuramoto reservoir has a universal approximation property. Furthermore, the conjecture on {\it the edge of bifurcation}, which is a generalization of the famous criterion {\it the edge of chaos} for designing a high-performance reservoir, is also discussed from the viewpoint of its approximation ability. It is numerically demonstrated by a prediction task and a transformation task.
シナプススケーリング則の最適化,モジュラスパイキングニューラルネットワークへの適用とその影響の解析

新川拓海, 加藤秀行, 石川慶孝, 住拓磨, 山本英明, 香取勇一

電子情報通信学会技術研究報告(Web)　123　(354(NLP2023 82-121))　2024年

ISSN： 2432-6380
レーザーマイクロダイセクションによるマイクロパターン培養神経回路への局所損傷

渡邉啓太, 渡邉啓太, 山本英明, 山本英明, 山本英明, 室田白馬, 室田白馬, 住拓磨, 佐藤茂雄, 佐藤茂雄, 平野愛弓, 平野愛弓, 平野愛弓

応用物理学会秋季学術講演会講演予稿集(CD-ROM)　85th　2024年

ISSN： 2758-4704
培養神経回路のネットワーク構造とリザバーコンピューティング招待有り

住拓磨, 山本英明, 山本英明, 千葉逸人, 香取勇一, 平野愛弓, 平野愛弓

日本神経回路学会誌　31　(3)　2024年

ISSN： 1340-766X
Modular Topology Enhances Reservoir Computing Performance in Biological Neuronal Networks 査読有り

Sumi Takuma, Yamamoto Hideaki, Katori Yuichi, Ito Koki, Sato Shigeo, Hirano-Iwata Ayumi

IEICE Proceeding Series　76　687-688　2023年9月21日
出版者・発行元： The Institute of Electronics, Information and Communication Engineers
DOI： 10.34385/proc.76.d2l-11 　

eISSN： 2188-5079

詳細を見る詳細を閉じる

Reservoir computing is a machine learning paradigm that employs high-dimensional dynamical systems for information processing. Although biological neuronal networks (BNNs) have been utilized to implement reservoir computing to provide insight into their computation properties, the neurons in conventional cultured neuronal networks are randomly connected, generating atypical dynamics characterized by highly correlated bursting activity not observed in healthy brains. In this study, we used micropatterning technology to fabricate BNNs with modular topology, a structural feature conserved in brain networks, and to understand how the dynamics within non-random networks of neuronal cells are linked to computing. Our study demonstrated that the modular BNN reservoir is capable of classifying both image and time-series data above chance levels. The modular structure in BNN contributes to the increased reservoir computing performance, in line with previous computational models with neuromorphic networks. Combining experiments with biological neuronal network and computational modeling can advance our understanding of computing principles in multicellular neuronal networks.
Evaluation of Reservoir-Based Predictive Coding in Cultured Neurons with Spiking Neural Network Model 査読有り

Ishikawa Yoshitaka, Shinkawa Takumi, Sumi Takuma, Kato Hideyuki, Yamamoto Hideaki, Katori Yuichi

IEICE Proceeding Series　76　693-696　2023年9月21日
出版者・発行元： The Institute of Electronics, Information and Communication Engineers
DOI： 10.34385/proc.76.d2l-14 　

eISSN： 2188-5079

詳細を見る詳細を閉じる

The field of physical reservoir computing, an approach that leverages the dynamics of physical systems for information processing, continues to evolve. Utilizing cultured neurons as a physical system in physical reservoir computing has helped to elucidate the relationship between the structure of neuronal networks and information processing. In this study, we present a model that integrates predictive coding and reservoir computing in cultured neurons, serving as a sensory information processing model. We demonstrate that the relationship between the structure of neuronal networks and the number of neurons influences sensory information processing. This study may provide a foundation for further investigations into information processing in biological neuronal networks.
Spiking and Bursting Properties of Spontaneous Activities in Modular Spiking Neuronal Networks 査読有り

Kato Hideyuki, Shinkawa Takumi, Ishikawa Yoshitaka, Sumi Takuma, Yamamoto Hideaki, Katori Yuichi

IEICE Proceeding Series　76　601-601　2023年9月21日
出版者・発行元： The Institute of Electronics, Information and Communication Engineers
DOI： 10.34385/proc.76.c4l-15 　

eISSN： 2188-5079

詳細を見る詳細を閉じる

The application of cultured neuronal networks as a physical reservoir is one of the attractive topics in reservoir computing. One can easily imagine that the performance of reservoir computing depends on their network topology because the topology influences neuronal activities in the network. In contrast, neurons try to sustain their activities as possible. This property is called the homeostasis of neurons. In this study, we mathematically model the networks with synaptically connected spiking neurons by considering the above two aspects and then evaluate spontaneous neuronal activities from the viewpoints of spiking and bursting behaviors.
Bottom-Up Investigation of Multicellular Computing Within Biological Neuronal Networks 査読有り

Yamamoto Hideaki, Sumi Takuma, Sato Yuya, Sato Shigeo, Hirano-Iwata Ayumi

IEICE Proceeding Series　76　594-595　2023年9月21日
出版者・発行元： The Institute of Electronics, Information and Communication Engineers
DOI： 10.34385/proc.76.c4l-11 　

eISSN： 2188-5079

詳細を見る詳細を閉じる

In this presentation, we will first introduce our studies aimed at reproducing an evolutionarily-conserved net-work structure in cultured neuronal networks on engineered glass coverslips and CMOS-based high-density microelectrode arrays. We then describe our recent attempts to couple the engineered neuronal networks with external stimulation to reveal their response to noise and spatiotemporally-patterned inputs to show that bioengineering technologies offer novel tools in investigating computational aspects of multicellular networks of biological neurons.
スパイキングニューロンモデルによる培養神経系のリザバー計算の性能評価

石川慶孝, 新川拓海, 住拓磨, 加藤秀行, 山本英明, 香取勇一

電子情報通信学会技術研究報告(Web)　122　(373(NLP2022 81-106))　2023年

ISSN： 2432-6380
大規模モジュール構造型培養神経回路のリザバー特性の解析

伊藤亘輝, 伊藤亘輝, 山本英明, 住拓磨, 住拓磨, 香取勇一, 佐藤茂雄, 平野愛弓, 平野愛弓, 平野愛弓

応用物理学会秋季学術講演会講演予稿集(CD-ROM)　84th　2023年

ISSN： 2758-4704
モジュール構造型培養神経回路のリザバー計算特性の解析

住拓磨, 住拓磨, 山本英明, 伊藤亘輝, 伊藤亘輝, 香取勇一, 佐藤茂雄, 平野愛弓, 平野愛弓, 平野愛弓

応用物理学会春季学術講演会講演予稿集(CD-ROM)　70th　2023年

ISSN： 2758-4704
大規模モジュール構造型培養神経回路の作製と自発活動の解析

伊藤亘輝, 伊藤亘輝, 山本英明, 山本英明, 山本英明, 住拓磨, 住拓磨, 室田白馬, 室田白馬, 門間信明, 門間信明, 佐藤茂雄, 佐藤茂雄, 佐藤茂雄, 平野愛弓, 平野愛弓, 平野愛弓, 平野愛弓, 平野愛弓

応用物理学会春季学術講演会講演予稿集(CD-ROM)　70th　2023年

ISSN： 2758-4704
人工神経細胞回路を用いた物理リザバーコンピューティング

山本英明, 住拓磨, 平野愛弓, 平野愛弓, 佐藤茂雄

化学とマイクロ・ナノシステム学会研究会講演要旨集(CD-ROM)　48th　2023年
Time-Series Classification in Micropatterned Neuronal Network Reservoirs 査読有り

Takuma Sumi, Hideaki Yamamoto, Yuichi Katori, Koki Ito, Shigeo Sato, Ayumi Hirano-Iwata

IEICE Proceeding Series　71　173-175　2022年12月12日
出版者・発行元： The Institute of Electronics, Information and Communication Engineers
DOI： 10.34385/proc.71.a5l-d-02 　

eISSN： 2188-5079

詳細を見る詳細を閉じる

Reservoir computing provides a novel framework to understand how the dynamics within biological neuronal networks (BNNs) is linked to information processing. Here, we used micropatterned substrates to fabricate BNNs with modular topology, one of the important structural features of brain networks, and realized a reservoir system with the modular BNN. Using image and time-series classification tasks, we evaluated the reservoir computing properties of the BNN reservoirs. The results show that modularity facilitates the separation between the trajectories of the neuronal responses to different spatial patterns, pointing to the functional advantage of the animals to modular topology within the nervous systems.
神経アンサンブルの機能特性解析のための人工神経細胞回路の構築

室田白馬, 室田白馬, 山本英明, 山本英明, 住拓磨, 住拓磨, 佐藤茂雄, 佐藤茂雄, 平野愛弓, 平野愛弓, 平野愛弓, 平野愛弓

電子情報通信学会技術研究報告(Web)　122　(195(NC2022 32-42))　2022年

ISSN： 2432-6380
モジュール構造型人工神経細胞回路の摂動応答特性

山本英明, 竹室汰貴, 竹室汰貴, 住拓磨, 住拓磨, 金野智浩, SORIANO Jordi, 佐藤茂雄, 平野愛弓, 平野愛弓, 平野愛弓

応用物理学会秋季学術講演会講演予稿集(CD-ROM)　83rd　2022年

ISSN： 2758-4704
In silico/in vitro神経回路におけるリザーバ計算特性のモジュール構造依存性

住拓磨, 山本英明, 竹室汰貴, 守谷哲, 佐藤茂雄, 平野愛弓

応用物理学会学術講演会講演予稿集　2021.1　2185-2185　2021年2月26日
出版者・発行元：公益社団法人応用物理学会
DOI： 10.11470/jsapmeeting.2021.1.0_2185 　

ISSN： 2758-4704

eISSN： 2436-7613
マイクロパターン培養神経回路のリザバー計算特性の解析

住拓磨, 住拓磨, 山本英明, 守谷哲, 竹室汰貴, 竹室汰貴, 金野智浩, 佐藤茂雄, 平野愛弓, 平野愛弓, 平野愛弓

電子情報通信学会技術研究報告(Web)　121　(223(NC2021 18-27))　2021年

ISSN： 2432-6380
リザバー計算に基づく人工神経細胞回路のパターン分類特性の評価

住拓磨, 住拓磨, 山本英明, 守谷哲, 竹室汰貴, 竹室汰貴, 金野智浩, 佐藤茂雄, 平野愛弓, 平野愛弓, 平野愛弓

日本神経回路学会全国大会講演論文集　31st (CD-ROM)　2021年
マイクロパターン培養神経回路に対する光摂動系の構築

竹室汰貴, 竹室汰貴, 山本英明, 脇村桂, 脇村桂, 住拓磨, 住拓磨, 金野智浩, 佐藤茂雄, SORIANO Jordi, 平野愛弓, 平野愛弓, 平野愛弓

応用物理学会秋季学術講演会講演予稿集(CD-ROM)　82nd　2021年

ISSN： 2758-4704
多点電極アレイ上でのモジュール構造型培養神経回路のパターニング

佐藤有弥, 佐藤有弥, 山本英明, 山本英明, 竹室汰貴, 竹室汰貴, 住拓磨, 住拓磨, 酒井原一守, 酒井原一守, 谷井孝至, 佐藤茂雄, 佐藤茂雄, 平野愛弓, 平野愛弓, 平野愛弓, 平野愛弓

応用物理学会秋季学術講演会講演予稿集(CD-ROM)　82nd　2021年

ISSN： 2758-4704
マイクロ加工培養神経回路の外部ノイズ刺激応答の解析

住拓磨, 山本英明, 脇村桂, 佐藤茂雄, 平野愛弓

応用物理学会学術講演会講演予稿集　2020.1　3653-3653　2020年2月28日
出版者・発行元：公益社団法人応用物理学会
DOI： 10.11470/jsapmeeting.2020.1.0_3653 　

ISSN： 2758-4704

eISSN： 2436-7613
脳組織に近い弾性率を有するシリコーン樹脂の生体界面材料応用—応用物理学会有機分子・バイオエレクトロニクス分科会研究会ここまで進んだ有機分子・バイオエレクトロニクス研究

山本英明, 住拓磨, 佐藤茂雄, 平野愛弓

Molecular electronics and bioelectronics = 応用物理学会,有機分子・バイオエレクトロニクス分科会会誌 / 応用物理学会有機分子・バイオエレクトロニクス分科会編　31　(2)　77-80　2020年
出版者・発行元：応用物理学会有機分子・バイオエレクトロニクス分科会
ISSN： 2423-8805
生体脳組織の弾性率を模倣した超軟シリコーン樹脂上での神経細胞培養

住拓磨, 山本英明, 平野愛弓

応用物理学会学術講演会講演予稿集　2019.2　2852-2852　2019年9月4日
出版者・発行元：公益社団法人応用物理学会
DOI： 10.11470/jsapmeeting.2019.2.0_2852 　

ISSN： 2758-4704

eISSN： 2436-7613
超軟ゲル材料表面を用いた神経細胞の生体模倣培養技術の開発

住拓磨, 山本英明, 早川岳志, 井手克哉, 木野久志, 田中徹, 平野愛弓

表面科学学術講演会要旨集　2018　180　2018年
出版者・発行元：公益社団法人日本表面真空学会
DOI： 10.14886/sssj2008.2018.0_180 　

詳細を見る詳細を閉じる

近年，細胞のメカノセンシティビティに関する研究が多くされており，細胞の振る舞いと細胞周辺の物理量の関係性が注目されている．我々は，基質材料の弾性率と神経細胞のシナプス特性との関係性に注目し，脳の弾性率を模倣した超軟ゲル材料上でラット大脳皮質神経細胞を培養する手法を確立し，パッチクランプ法およびCaイメージングにより神経活動を記録，解析したので，その結果を報告する．

︎全件表示 ︎最初の5件までを表示