13197 research outputs found
Sort by
Millimeter-Wave Band Electro-Optical Imaging System Using Polarization CMOS Image Sensor and Amplified Optical Local Oscillator Source
No full textIn this study, we developed and demonstrated a millimeter-wave electric field imaging system using an electro-optic crystal and a highly sensitive polarization measurement technique using a polarization image sensor, which was fabricated using a 0.35-00B5m at a rate of 1 FPS, corresponding to 2400 pixels/s.journal articl
Brain-implantable needle-type CMOS imaging device enables multi-layer dissection of seizure calcium dynamics in the hippocampus
No full textObjective: Current neuronal imaging methods mostly use bulky lenses that either impede animal behavior or prohibit multi-depth imaging. To overcome these limitations, we developed a lightweight lensless biophotonic system for neuronal imaging, enabling compact and simultaneous visualization of multiple brain layers. Approach: Our developed 'CIS-NAIST' device integrates a micro-CMOS image sensor, thin-film fluorescence filter, micro-LEDs, and a needle-shaped flexible printed circuit. With this device, we monitored neuronal calcium dynamics during seizures across the different layers of the hippocampus and employed machine learning techniques for seizure classification and prediction. Main results: The CIS-NAIST device revealed distinct calcium activity patterns across the CA1, molecular interlayer, and dentate gyrus. Our findings indicated an elevated calcium amplitude activity specifically in the dentate gyrus compared to other layers. Then, leveraging the multi-layer data obtained from the device, we successfully classified seizure calcium activity and predicted seizure behavior using Long Short-Term Memory and Hidden Markov models. Significance: Taken together, our 'CIS-NAIST' device offers an effective and minimally invasive method of seizure monitoring that can help elucidate the mechanisms of temporal lobe epilepsy.journal articl
Generation of rat-derived lung epithelial cells in Fgfr2b-deficient mice retains species-specific development
Get PDFRegenerative medicine is a tool to compensate for the shortage of lungs for transplantation, but it remains difficult to construct a lung in vitro due to the complex three-dimensional structures and multiple cell types required. A blastocyst complementation method using interspecies chimeric animals has been attracting attention as a way to create complex organs in animals, although successful lung formation using interspecies chimeric animals has not yet been achieved. Here, we applied a reverse-blastocyst complementation method to clarify the conditions required to form lungs in an Fgfr2b-deficient mouse model. We then successfully formed a rat-derived lung in the mouse model by applying a tetraploid-based organ-complementation method. Importantly, rat lung epithelial cells retained their developmental timing even in the mouse body. These findings provide useful insights to overcome the barrier of species-specific developmental timing to generate functional lungs in interspecies chimeras.journal articl
Linking Code and Documentation Churn: Preliminary Analysis
No full textCode churn refers to the measure of the amount of code added, modified, or deleted in a project and is often used to assess codebase stability and maintainability. Program comprehension or how understandable the changes are, is equally important for maintainability. Documentation is crucial for knowledge transfer, especially when new maintainers take over abandoned code. We emphasize the need for corresponding documentation updates, as this reflects project health and trustworthiness as a third-party library. Therefore, we argue that every code change should prompt a documentation update (defined as documentation churn). Linking code churn changes with documentation updates is important for project sustainability, as it facilitates knowledge transfer and reduces the effort required for program comprehension. This study investigates the synchrony between code churn and documentation updates in three GitHub open-source projects. We will use qualitative analysis and repository mining to examine the alignment and correlation of code churn and documentation updates over time. We want to identify which code changes are likely synchronized with documentation and to what extent documentation can be auto-generated. Preliminary results indicate varying degrees of synchrony across projects, highlighting the importance of integrated concurrent documentation practices and providing insights into how recent technologies like AI, in the form of Large Language Models (i.e., LLMs), could be leveraged to keep code and documentation churn in sync. The novelty of this study lies in demonstrating how synchronizing code changes with documentation updates can improve the development lifecycle by enhancing diversity and efficiency.conference pape
Cross-lingual Natural Language Processing on Limited Annotated Case/Radiology Reports in English and Japanese: Insights from the Real-MedNLP Workshop
No full textBackground2003We present the outcomes of the Real-MedNLP workshop, which was conducted using limited and parallel medical corpora. Real-MedNLP exhibits three distinct characteristics: (1) limited annotated documents: the training data comprise only a small set (2003We propose three tasks: NER of 2003Nine teams participated in this study. The best systems achieved 0.65 and 0.89 F1-scores for CRs and RRs in Task 1, whereas the top scores in Task 2 decreased by 50 to 70%. In Task 3, ADE reports were detected by up to 0.64 F1-score, and CI scored up to 0.96 binary accuracy. Conclusion2013specific pretrained language models using data augmentation methods. Despite the challenge of limited corpus size in Tasks 1 and 2, recent approaches are promising because the partial match scores reached 20130.9$2009F1-scores. Task 3 applications revealed that the different availabilities of external language resources affected the performance per language.journal articl
Astrocyte-Specific Inhibition of the Primary Cilium Suppresses C3 Expression in Reactive Astrocyte
No full textC3-positive reactive astrocytes play a neurotoxic role in various neurodegenerative diseases. However, the mechanisms controlling C3-positive reactive astrocyte induction are largely unknown. We found that the length of the primary cilium, a cellular organelle that receives extracellular signals was increased in C3-positive reactive astrocytes, and the loss or shortening of primary cilium decreased the count of C3-positive reactive astrocytes. Pharmacological experiments suggested that Ca2+ signalling may synergistically promote C3 expression in reactive astrocytes. Conditional knockout (cKO) mice that specifcally inhibit primary cilium formation in astrocytes upon drug stimulation exhibited a reduction in the proportions of C3-positive reactive astrocytes and apoptotic cells in the brain even after the injection of lipopolysaccharide (LPS). Additionally, the novel object recognition (NOR) score observed in the cKO mice was higher than that observed in the neuroinfammation model mice. These results suggest that the primary cilium in astrocytes positively regulates C3 expression. We propose that regulating astrocyte-specifc primary cilium signalling may be a novel strategy for the suppression of neuroinfammation.journal articl
Agarose Nanofiber by Electrospinning
Get PDFAgarose nanofiber with a diameter of 68 ± 33 nm is first prepared by electrospinning under an optimized condition of solvents with hexafluoroisopropanol/water (92.5/7.5, v/v). The results show the control of hydrogen bonding, which is important to prepare a nanoscale agarose material.journal articl
Multi-label Learning with Random Circular Vectors
No full textThe extreme multi-label classification (XMC) task involves learning a classifier that can predict from a large label set the most relevant subset of labels for a data instance. While deep neural networks (DNNs) have demonstrated remarkable success in XMC problems, the task is still challenging because it must deal with a large number of output labels, which make the DNN training computationally expensive. This paper addresses the issue by exploring the use of random circular vectors, where each vector component is represented as a complex amplitude. In our framework, we can develop an output layer and loss function of DNNs for XMC by representing the final output layer as a fully connected layer that directly predicts a low-dimensional circular vector encoding a set of labels for a data instance. We conducted experiments on synthetic datasets to verify that circular vectors have better label encoding capacity and retrieval ability than normal real-valued vectors. Then, we conducted experiments on actual XMC datasets and found that these appealing properties of circular vectors contribute to significant improvements in task performance compared with a previous model using random real-valued vectors, while reducing the size of the output layers by up to 99%.conference pape
ノリモノ ヨイ ケイゲン オ モクテキ ト シタ イロ ト カイゾウド ノ セイギョ ニ モトズク ムイシキテキ ナ シセン ユウドウ システム
Get PDF視線動向は乗り物酔いの発生原因に密接に関係していると考えられており、視線誘導を用いた視線動向の改善によって、乗り物酔いの軽減効果が期待されている。従来研究では、明示的な視覚刺激を与える視線誘導が採用されているため、ユーザエクスペリエンスを低減させるという課題が存在した。本研究では、明示的な視覚刺激によらない無意識的な視線誘導法、具体的には提示映像の解像度および色の制御による視線誘導法を提案し、車窓映像における有効性を検証する。また、二つ（解像度と色）の手法を組み合わせることで、より強い視線誘導の可能性についても検証する。conference pape
イリョウ エーアイ ガ カノウ ニ スル ジセダイ ノ セイシン イリョウ ― エーアイ ガ ヒラク ジセダイ ノ セイシン イリョウ ―
No full text今，医療が変わりつつある．電子カルテに集積される医療ビッグデータ，それを用いた人工知能による診断支援，さらには，スマートフォンやスマートスピーカといった新たなデバイスからの時系列情報など，さまざまな材料，技術が登場しており，膨大な情報が利用可能になりつつある．特に，大規模言語モデル，いわゆるAIが，これまで扱いが困難であった自然言語をある程度処理できるようになったことのインパクトは大きく，対話インタフェースを用いた多くのアプリケーションが開発されつつある．この結果，医療者のみならず，患者側も主体的にAI技術を使い始め，医療に関する情報収集，カウンセリング，行動変容など医療とそれを取り巻く周辺で利用されつつある．この流れはもう止められるものではなく，今後さらに多くの患者がAIを利用するのは間違いないであろう．楽観的には，AIが患者と医療者の両方をサポートし，連携を深めることで，患者に寄り添った緻密な医療が実現する可能性もある．しかし，AIがブラックボックスで原理についてまだ不明な点が多いこと，機械の説明責任，医療の信頼，診断の公平性，データの所有権など，多くの問題がまだ十分に議論されないままであり，AIが予想もしない問題を引き起こしてしまう可能性も危惧される．本稿では，医療AIの現況を俯瞰し，次世代の精神医療に起こり得るかもしれない問題について議論する．journal articl