Yukako Tohsato
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Yukako Tohsato has a total of 40 co-authors. Here are the top co-authors by number of shared publications:
- Koji Kyoda
RIKEN Center for Biosystems Dynamics Research
4 shared publications - Shuichi Onami
Laboratory for Developmental Dynamics, RIKEN Center for Biosystems Dynamics Research
4 shared publications - Hiroya Itoga
RIKEN Center for Biosystems Dynamics Research
3 shared publications - Ikuko Nishikawa
Ritsumeikan University
3 shared publications - Takaya Ueda
Ritsumeikan University
3 shared publications - Kenneth H. L. Ho
Crick Advanced Light Microscopy Facility The Francis Crick Institute London UK
2 shared publications
Latest Publications
- Open Access
- 5 November 2026
- Published by MDPI AG in Applied Sciences
- Vol. 16 (21) , 10953
Robust Outside-Range Optimization by Decoupling Search-Region Expansion and Search-Distribution Update in AREX/JGG
In real-coded genetic algorithms (RCGAs), search performance often deteriorates when the true global optimum lies outside the initial search range because crossover-based sampling remains concentrated around the initial population. We define this problem as outside-range optimization and introduce the outside-range ratio, representing the proportion of dimensions whose true optimum lies outside the initial search range. We extend Adaptive Real-coded Ensemble Crossover/Just Generation Gap (AREX/JGG) by explicitly decoupling search-region expansion and search-distribution update. Specifically, enhanced Chaotic Search (e-CS) expands the search region through extrapolation, AREX updates the search distribution, and Combinational Mutation (CM) generates candidates with heavy-tailed perturbations. Experiments on 10 benchmark functions with 20–80 dimensions showed lower expected running time (ERT) than AREX/JGG under many outside-range conditions examined. For 40-dimensional Sphere, Cigar, and k-tablet, ERT was lower than HIRCGA across all tested outside-range ratios and distances. Furthermore, for parameter estimation of an ordinary differential equation (ODE) model of Escherichia coli glycolysis, the proposed method achieved a higher success rate than AREX/JGG in three of four initial-range settings. These results suggest that the proposed method can improve search efficiency under many tested outside-range conditions and enhance the robustness of search success to initial-range settings.
- 5 November 2026
- Published by MDPI AG in Applied Sciences
- Vol. 16 (21) , 10953
Open AccessRobust Outside-Range Optimization by Decoupling Search-Region Expansion and Search-Distribution Update in AREX/JGG
In real-coded genetic algorithms (RCGAs), search performance often deteriorates when the true global optimum lies outside the initial search range because crossover-based sampling remains concentrated around the initial population. We define this problem as outside-range optimization and introduce the outside-range ratio, representing the proportion of dimensions whose true optimum lies outside the initial search range. We extend Adaptive Real-coded Ensemble Crossover/Just Generation Gap (AREX/JGG) by explicitly decoupling search-region expansion and search-distribution update. Specifically, enhanced Chaotic Search (e-CS) expands the search region through extrapolation, AREX updates the search distribution, and Combinational Mutation (CM) generates candidates with heavy-tailed perturbations. Experiments on 10 benchmark functions with 20–80 dimensions showed lower expected running time (ERT) than AREX/JGG under many outside-range conditions examined. For 40-dimensional Sphere, Cigar, and k-tablet, ERT was lower than HIRCGA across all tested outside-range ratios and distances. Furthermore, for parameter estimation of an ordinary differential equation (ODE) model of Escherichia coli glycolysis, the proposed method achieved a higher success rate than AREX/JGG in three of four initial-range settings. These results suggest that the proposed method can improve search efficiency under many tested outside-range conditions and enhance the robustness of search success to initial-range settings.
- Open Access
Phenotype-driven screening reveals a causal role for the cortex in pupil control
Summary: Phenotype-driven forward screening offers a powerful strategy to discover neuronal substrates underlying physical and behavioral traits without prior anatomical or functional assumptions. Historically, such approaches have identified many important neuronal circuits using invertebrate model organisms, but application to mammals has remained limited by the lack of appropriate strategies. Here we quantitatively profiled 56 neurological phenotypic features across more than 200 adult mice when neurodevelopmentally classified neurons were activated or inhibited by chemogenetics. This screen yielded a wide spectrum of robust neurological phenotypes. Upon analyzing these phenotypes computationally, we narrowed down to the hypothesis that activation of cortical neurons enlarges pupil size. Experimental evidence using optogenetics and in utero electroporation indicated that this hypothesis is true. Our results provide proof of principle that phenotype-driven forward approach in mammals is a powerful alternative as a laboratory approach to uncover brain substrates, and offers a general framework for systematically mapping neural circuits that regulate physical and behavioral phenotypes.
Open AccessPhenotype-driven screening reveals a causal role for the cortex in pupil control
Summary: Phenotype-driven forward screening offers a powerful strategy to discover neuronal substrates underlying physical and behavioral traits without prior anatomical or functional assumptions. Historically, such approaches have identified many important neuronal circuits using invertebrate model organisms, but application to mammals has remained limited by the lack of appropriate strategies. Here we quantitatively profiled 56 neurological phenotypic features across more than 200 adult mice when neurodevelopmentally classified neurons were activated or inhibited by chemogenetics. This screen yielded a wide spectrum of robust neurological phenotypes. Upon analyzing these phenotypes computationally, we narrowed down to the hypothesis that activation of cortical neurons enlarges pupil size. Experimental evidence using optogenetics and in utero electroporation indicated that this hypothesis is true. Our results provide proof of principle that phenotype-driven forward approach in mammals is a powerful alternative as a laboratory approach to uncover brain substrates, and offers a general framework for systematically mapping neural circuits that regulate physical and behavioral phenotypes.
- 12 January 2026
- Published by Springer Nature
- p. 52-66
SAPAG: Self-attention Parts Guidance with Latent Diffusion Models for Anomaly Detection in Microscopy Images
- 12 January 2026
- Published by Springer Nature
- p. 52-66
SAPAG: Self-attention Parts Guidance with Latent Diffusion Models for Anomaly Detection in Microscopy Images
- Open Access
- 1 December 2025
- Published by Elsevier in Journal of Hazardous Materials
- Vol. 501, 140678
Suburban-urban differences in coarse and fine atmospheric particulate matter with key chemical compositions influencing bacterial communities in Toyama and Yokohama, Japan
- 1 December 2025
- Published by Elsevier in Journal of Hazardous Materials
- Vol. 501, 140678
Open AccessSuburban-urban differences in coarse and fine atmospheric particulate matter with key chemical compositions influencing bacterial communities in Toyama and Yokohama, Japan
- Open Access
- 31 October 2024
- Published by Oxford University Press (OUP) in Nucleic Acids Research
- Vol. 53 (D1) , D1716-D1723
SSBD: an ecosystem for enhanced sharing and reuse of bioimaging data
SSBD (https://ssbd.riken.jp) is a platform for the sharing and reuse of bioimaging data. As part of efforts to build a bioimaging data ecosystem, SSBD has recently been updated to a two-tiered data resource comprising SSBD:repository, a public repository for the sharing of all types of bioimaging data reported in journals, and SSBD:database, an added-value database for the sharing of curated, highly reusable, metadata-rich data. This update addresses the conflicting demands of rapid data publication and sharing of richly annotated data, thereby promoting bioimaging data sharing and reuse. With this update, SSBD is now positioned as a core repository and database within the foundingGIDE, an international consortium working to establish a global image data ecosystem. Harmonizing metadata between data resources enables cross-searching and data exchange with data resources from other countries and regions.
- 31 October 2024
- Published by Oxford University Press (OUP) in Nucleic Acids Research
- Vol. 53 (D1) , D1716-D1723
Open AccessSSBD: an ecosystem for enhanced sharing and reuse of bioimaging data
SSBD (https://ssbd.riken.jp) is a platform for the sharing and reuse of bioimaging data. As part of efforts to build a bioimaging data ecosystem, SSBD has recently been updated to a two-tiered data resource comprising SSBD:repository, a public repository for the sharing of all types of bioimaging data reported in journals, and SSBD:database, an added-value database for the sharing of curated, highly reusable, metadata-rich data. This update addresses the conflicting demands of rapid data publication and sharing of richly annotated data, thereby promoting bioimaging data sharing and reuse. With this update, SSBD is now positioned as a core repository and database within the foundingGIDE, an international consortium working to establish a global image data ecosystem. Harmonizing metadata between data resources enables cross-searching and data exchange with data resources from other countries and regions.
- Open Access
- 1 January 2024
- Published by Information Processing Society of Japan in IPSJ Transactions on Bioinformatics
- Vol. 17, 33-39
Segmentation of Mouse Brain Slices with Unsupervised Domain Adaptation Considering Cross-sectional Locations
Images of mouse brain slices, obtained under slightly different experimental conditions, are available in 84 datasets in the NeuroGT database (https://ssbd.riken.jp/neurogt/). Our goal was to obtain semantic segmentation results for eight brain anatomical regions. However, out of 84 datasets, only one dataset had true labels that could be used to train a convolutional neural network (CNN), and it was incomplete (131 out of 162 images). A segmentation model trained with the labeled images was less accurate on other images obtained under different experimental conditions because of differences of the image properties. We therefore tried Unsupervised Domain Adaptation (UDA), wherein the parameters of the CNN trained on the labeled images (source) were transferred to the unlabeled images (target). We used the positional information of the sample slices associated with each image to propose a novel loss function that approximated the class occurrence probabilities of segmentation results obtained from source and target images of brain samples at similar sliced locations, and we introduced it into the UDA. The proposed UDA method achieved an mIoU of 78.34%, which was 8% more accurate than the previous UDA methods such as Contrastive Learning and Self-Training (CLST) and Maximum Classifier Discrepancy (MCD). We demonstrated experimentally that the proposed method was useful for segmenting biomedical images with a small amount of incomplete training data.
- 1 January 2024
- Published by Information Processing Society of Japan in IPSJ Transactions on Bioinformatics
- Vol. 17, 33-39
Open AccessSegmentation of Mouse Brain Slices with Unsupervised Domain Adaptation Considering Cross-sectional Locations
Images of mouse brain slices, obtained under slightly different experimental conditions, are available in 84 datasets in the NeuroGT database (https://ssbd.riken.jp/neurogt/). Our goal was to obtain semantic segmentation results for eight brain anatomical regions. However, out of 84 datasets, only one dataset had true labels that could be used to train a convolutional neural network (CNN), and it was incomplete (131 out of 162 images). A segmentation model trained with the labeled images was less accurate on other images obtained under different experimental conditions because of differences of the image properties. We therefore tried Unsupervised Domain Adaptation (UDA), wherein the parameters of the CNN trained on the labeled images (source) were transferred to the unlabeled images (target). We used the positional information of the sample slices associated with each image to propose a novel loss function that approximated the class occurrence probabilities of segmentation results obtained from source and target images of brain samples at similar sliced locations, and we introduced it into the UDA. The proposed UDA method achieved an mIoU of 78.34%, which was 8% more accurate than the previous UDA methods such as Contrastive Learning and Self-Training (CLST) and Maximum Classifier Discrepancy (MCD). We demonstrated experimentally that the proposed method was useful for segmenting biomedical images with a small amount of incomplete training data.
- Open Access
- 15 June 2023
- Published by Association for Computing Machinery (ACM)
- p. 1-7
Large-scale Image Processing and Three-Dimensional Reconstruction of Mouse Brains with Neurogenic-Tagged Neurons
The NeuroGT (neurogenic tagging) database provides images of whole-brain sections of mice in which neurons that differentiated at a similar time have been genetically tagged. The images of brain sections are displayed in order from the anterior to the posterior end on a web-browser, and the regions containing tagged neurons can be visually confirmed. However, it has been difficult to grasp their location in three-dimensional (3D) space. As a first step toward quantitative analysis of neuronal location, we automatically detected regions of mouse brains containing tagged neurons by using image processing and integrated that information into 3D space. We confirmed that the proposed methods detected brain regions with an average precision of 88% and average recall of 99%. We achieved visualization of 3D reconstructed brain shapes and distribution of tagged neurons in the regions of the brain.
- 15 June 2023
- Published by Association for Computing Machinery (ACM)
- p. 1-7
Open AccessLarge-scale Image Processing and Three-Dimensional Reconstruction of Mouse Brains with Neurogenic-Tagged Neurons
The NeuroGT (neurogenic tagging) database provides images of whole-brain sections of mice in which neurons that differentiated at a similar time have been genetically tagged. The images of brain sections are displayed in order from the anterior to the posterior end on a web-browser, and the regions containing tagged neurons can be visually confirmed. However, it has been difficult to grasp their location in three-dimensional (3D) space. As a first step toward quantitative analysis of neuronal location, we automatically detected regions of mouse brains containing tagged neurons by using image processing and integrated that information into 3D space. We confirmed that the proposed methods detected brain regions with an average precision of 88% and average recall of 99%. We achieved visualization of 3D reconstructed brain shapes and distribution of tagged neurons in the regions of the brain.
- Open Access
- 1 January 2023
- Published by Information Processing Society of Japan in IPSJ Transactions on Bioinformatics
- Vol. 16, 13-19
Metabolic Network Analysis by Time-series Causal Inference Using the Multi-dimensional Space of Prediction Errors
Identifying factors that contribute to microbial growth is important for realizing efficient production of useful substrates. Our objective was to predict unknown metabolic pathways from experimental time-series data in model organisms such as Escherichia coli. We focused on a previous method that replaces the computation of auto-regression in the Granger causality test with non-parametric multiplicative regression (NPMR) to allow inferences on noisy and nonlinear data. We then proposed a new causal inference method that creates a multi-dimensional space based on the error between the time series predicted by NPMR and the original time series. We confirmed that the inference accuracy of the proposed method outperforms that of NPMR by 50% using short time series generated by coupled logistic equations, which allows for adjustment of the strength of the causal relationship. The proposed method was applied to simulation data obtained from a kinetic model for glycolysis in E. coli and achieved 61% accuracy.
- 1 January 2023
- Published by Information Processing Society of Japan in IPSJ Transactions on Bioinformatics
- Vol. 16, 13-19
Open AccessMetabolic Network Analysis by Time-series Causal Inference Using the Multi-dimensional Space of Prediction Errors
Identifying factors that contribute to microbial growth is important for realizing efficient production of useful substrates. Our objective was to predict unknown metabolic pathways from experimental time-series data in model organisms such as Escherichia coli. We focused on a previous method that replaces the computation of auto-regression in the Granger causality test with non-parametric multiplicative regression (NPMR) to allow inferences on noisy and nonlinear data. We then proposed a new causal inference method that creates a multi-dimensional space based on the error between the time series predicted by NPMR and the original time series. We confirmed that the inference accuracy of the proposed method outperforms that of NPMR by 50% using short time series generated by coupled logistic equations, which allows for adjustment of the strength of the causal relationship. The proposed method was applied to simulation data obtained from a kinetic model for glycolysis in E. coli and achieved 61% accuracy.
- 15 September 2022
- Published by Springer Nature
- p. 469-481
Unsupervised Domain Adaptation Using Temporal Association for Segmentation and Its Application to C. elegans Time-Lapse Images
- 15 September 2022
- Published by Springer Nature
- p. 469-481
Unsupervised Domain Adaptation Using Temporal Association for Segmentation and Its Application to C. elegans Time-Lapse Images
- 7 September 2022
- Published by Springer Nature
- p. 432-444
Phenotype Anomaly Detection for Biological Dynamics Data Using a Deep Generative Model
- 7 September 2022
- Published by Springer Nature
- p. 432-444
Phenotype Anomaly Detection for Biological Dynamics Data Using a Deep Generative Model