A5101762099- Cell Image Analysis Techniques
- 3D Printing in Biomedical Research
- Single-cell and spatial transcriptomics
- Image Processing Techniques and Applications
- Semantic Web and Ontologies
- RNA Interference and Gene Delivery
- Advanced Data Compression Techniques
- Topic Modeling
- Generative Adversarial Networks and Image Synthesis
- CAR-T cell therapy research
- Multimodal Machine Learning Applications
- Extracellular vesicles in disease
- Computer Graphics and Visualization Techniques
- Image and Signal Denoising Methods
- Immune Cell Function and Interaction
- T-cell and B-cell Immunology
- Natural Language Processing Techniques
- MicroRNA in disease regulation
- Advanced Image Processing Techniques
- Advanced Optical Imaging Technologies
The University of Tokyo
2020-2025
The upgraded version of intelligent image-activated cell sorting (iIACS) has enabled higher-throughput and more sensitive image-based single live cells from heterogeneous populations.
Imaging flow cytometry (IFC) has become a powerful tool for diverse biomedical applications by virtue of its ability to image single cells in high-throughput manner. However, there remains challenge posed the fundamental trade-off between throughput, sensitivity, and spatial resolution. Here we present deep-learning-enhanced imaging (dIFC) that circumvents this implementing an restoration algorithm on virtual-freezing fluorescence (VIFFI) platform, enabling higher throughput without...
ABSTRACT T cells play a pivotal role in the immune system's response to various conditions. They are activated by antigen‐presenting (APCs) via T‐cell surface receptors, resulting cytokine production and proliferation. These interactions occur through formation of immunological synapses. The advent imaging flow cytometry has enabled detailed statistical analyses these cellular interactions. However, dynamics receptors vitro stimulation yet receive attention, despite it being crucial aspect...
Abstract Intelligent image‐activated cell sorting (iIACS) has enabled high‐throughput image‐based of single cells with artificial intelligence (AI) algorithms. This AI‐on‐a‐chip technology combines fluorescence microscopy, AI‐based image processing, sort‐timing prediction, and sorting. Sort‐timing prediction is particularly essential due to the latency on order milliseconds between acquisition sort actuation, during which processing performed. The long amplifies effects fluctuations in flow...
Homotypic targeting is the inherent ability of cells for preferential interaction with similar or identical types, a phenomenon commonly seen in cell adhesion, tissue formation, and immune responses. Unfortunately, its full potential remains largely untapped. Here we introduce an approach to drastically boost homotypic capabilities via exosomes (nanoscale extracellular vesicles secreted by cells). By engineering exosome surfaces lanthanides, amplify specific cell-exosome interactions more...
Does seeing always mean knowing? Large Vision-Language Models (LVLMs) integrate separately pre-trained vision and language components, often using CLIP-ViT as backbone. However, these models frequently encounter a core issue of "cognitive misalignment" between the encoder (VE) large model (LLM). Specifically, VE's representation visual information may not fully align with LLM's cognitive framework, leading to mismatch where features exceed model's interpretive range. To address this, we...
Achieving optimal performance of video diffusion transformers within given data and compute budget is crucial due to their high training costs. This necessitates precisely determining the model size hyperparameters before large-scale training. While scaling laws are employed in language models predict performance, existence accurate derivation visual generation remain underexplored. In this paper, we systematically analyze for confirm presence. Moreover, discover that, unlike models, more...
Multimodal Large Language Models (MLLMs) have recently demonstrated remarkable perceptual and reasoning abilities, typically comprising a Vision Encoder, an Adapter, Model (LLM). The adapter serves as the critical bridge between visual language components. However, training adapters with image-level supervision often results in significant misalignment, undermining LLMs' capabilities limiting potential of LLMs. To address this, we introduce Supervised Embedding Alignment (SEA), token-level...