13197 research outputs found
Sort by
Synthetic Utilization of 2H-Heptafluoropropane: Ionic 1,4-Addition to Electron-Deficient Carbon$2013Carbon Unsaturated Bonds
Get PDFWe have found a novel method for introducing heptafluoro-2-propyl CF(CF3)2 groups into carbon$2013carbon unsaturated bonds via a nucleophilic reaction using 2H-heptafluoropropane as the source of CF(CF3)2 groups. The reaction involves the nucleophilic addition of a heptafluoro-2-propyl anion, generated by treating 2H-heptafluoropropane with a fluoride ion, to various electron-deficient unsaturated compounds. This allows the easy synthesis of various aliphatic compounds containing heptafluoro-2-propyl groups.journal articl
Effect of Average Grain Size on the Uniformity and Ferroelectricity of BTO/PSX Thin Films Processed by Low-Temperature Solution Method
No full textIn this study, we utilized 50 nm BaTiO3 (BTO) nanoparticles and polysiloxane (PSX) with a higher concentration of methyl and silica groups to fabricate insulating layers at a low curing temperature of 100 °C using a solution-based method. This approach aims to enhance film uniformity while retaining the ferroelectric properties. Consequently, we maintained a minimal leakage current in thin-film transistors (TFTs) while achieving transfer characteristics characterized by a distinct hysteresis. Moreover, we verified the presence of ferroelectricity in 50 nm BTO nanoparticles. Compared with prior research, we confirm that decreasing nanoparticle size effectively reduces film roughness but also leads to a reduction in polarization intensity due to smaller diameter BTO nanoparticles. Additionally, a higher proportion of methyl and silica groups effectively lowers the curing temperature of PSX. At the same time, the hydrogen ions released in the polycondensation reaction can also effectively suppress the oxygen vacancies at the interface between dielectric and channel layers, improving the TFT electrical characteristics.journal articl
Preoperative Assessment of Vessel-to-acetabular Rim Distances in Non-contrast CT Images for Total Hip Arthroplasty
No full textPurpose: Blood vessel injuries during total hip arthroplasty (THA) pose life-threatening risks. This study aimed at creating a preoperative approach for evaluating vessel-to-acetabular rim distances for surgical planning of THA.
Methods: From non-contrast CT images, the pelvis and blood vessels including the external iliac artery and vein were automatically segmented. Surface points on the vessels and the acetabular rim were used to assess distance within high-risk locations. Thirty-six non-contrast CT images of patients with hip osteoarthritis (OA) and 18 pairs of contrast-enhanced/non-contrast CT images of patients with non-hip OA were used for internal training/validation, and 10 non-hip OA CT images collected from a public database were used for external validation. The ground truth (GT) labels were manually annotated for the internal hip OA database, whereas the arterial GT labels for the internal non-hip OA database were constructed using registered contrast-enhanced CT, in which the arteries were clearly imaged. The Dice coefficient (DC) and average symmetric surface distance (ASD) were used to evaluate segmentation accuracy. The accuracy of distance assessment was assessed using mean absolute error (MAE) and Pearson correlation coefficient (PCC).
Results: Using a 3D nnU-Net model, DCs for artery and vein were 0.877 and 0.892, respectively, in internal hip OA patients, 0.901 and 0.909 in internal non-hip OA patients, and 0.846 and 0.858 in external non-hip OA patients. PCCs between the GT and auto-segmentation-based distances were larger than 0.97, with a mean MAE smaller than 0.5 mm in the high-risk location.
Conclusions: The study proposes a preoperative approach for vessel-to-rim distance assessments in non-contrast CT images. The evaluation results demonstrated high accuracy in automated vessel segmentation and distance assessment, showing the feasibility of using the proposed method for risk assessment of vascular injury in THA surgical planning.journal articl
Graph Neural Networks for Personalized Recommendation
No full textvideo/mp4Graph Neural Networks (GNNs) have shown great success in recommender systems in recent years. In this presentation, I will first give a brief introduction of GNN and personalized recommendation followed by some successful instantiations of GNN-based recommendation methods. Then, I will also introduce some challenges and further potential of GNN-based recommendation.講演日: 2024年6月20日講演場所: エーアイ大講義室, AI Inc. Seminar Hall （L1）vide
Detection of multiple fluorescent dyes using liquid crystal variable retarder and sparse modeling
No full textIn spatial transcriptomics, which allows the analysis of gene expression while preserving its location in tissues, RNA molecules are hybridized with a fluorescent-labeled DNA probe for detection. In this study, we aim to improve the efficiency of spatial transcriptomics by simultaneously using multiple fluorescent dyes with overlapping spectra. We propose a method to quantify each fluorescent dyes using a liquid crystal variable retarder as a spectral modulator, which can control the spectral transmittance by changing the voltage. The spectrum of light passing through the modulator is integrated by the image sensor and observed as intensity. We quantify the fluorescent dyes at each pixel using intensities of various spectral transmittances as a spectral code and applying sparse modeling using a dictionary created by simulating observations for the fluorescent dyes used in hybridization. We verified the principle of the proposed method and demonstrated its feasibility through simulation experiments.conference pape
シナリオ ノ ドウニュウ ニ ヨル タヨウ ナ プログラミング エンシュウ モンダイ ノ ジドウ セイセイ シュホウ ノ ケントウ
Get PDFプログラミング演習では，学生が学習すべき内容に応じた演習問題を教員が大量に作成する必要がある．我々の研究グループでは，演習問題作成時に教員にかかる負荷の低減を目的として，自然な文章の自動生成を可能とする生成 AI を用いた自動生成システムを提案している．このシステムでは，演習問題ごとに題材や利用すべき文法要素等の生成条件を教員が入力すると，解答例コードと仕様，実行例が生成される．一方で，システムに入力する生成条件が類似している演習問題が複数あると，出力される解答例コードや仕様，実行例も類似してしまうという課題がある．実際のプログラミング演習で利用可能な演習問題を生成するには，同一の生成条件から複数の異なる多様な演習問題を生成可能であることが望ましい．そこで本研究では，シナリオというプログラムの流れを説明する短い文章を演習問題生成時に付与することで演習問題ごとのデータ構造やアルゴリズム，難易度が異なる多様な演習問題の生成を実現する手法を提案する．実際にシナリオを導入して演習問題を生成したところ，シナリオの長さや複雑さ，記述内容の差が生成される演習問題のデータ構造やアルゴリズム，難易度の差に影響を与える可能性があることが確認された．一方で作成される演習問題の難易度が上昇することで，演習問題の理解性が下がり，仕様や実行例を教員が修正しなければならないケースが増加する傾向が見られた．conference pape
An interpretable and transferrable vision transformer model for rapid materials spectra classification
Get PDFRapid analysis of materials characterization spectra is pivotal for preventing the accumulation of unwieldy datasets, thus accelerating subsequent decision-making. However, current methods heavily rely on experience and domain knowledge, which not only proves tedious but also makes it hard to keep up with the pace of data acquisition. In this context, we introduce a transferable Vision Transformer (ViT) model for the identification of materials from their spectra, including XRD and FTIR. First, an optimal ViT model was trained to predict metal organic frameworks (MOFs) from their XRD spectra. It attains prediction accuracies of 70%, 93%, and 94.9% for Top-1, Top-3, and Top-5, respectively, and a shorter training time of 269 seconds ($223C30% faster) in comparison to a convolutional neural network model. The dimension reduction and attention weight map underline its adeptness at capturing relevant features in the XRD spectra for determining the prediction outcome. Moreover, the model can be transferred to a new one for prediction of organic molecules from their FTIR spectra, attaining remarkable Top-1, Top-3, and Top-5 prediction accuracies of 84%, 94.1%, and 96.7%, respectively. The introduced ViT-based model would set a new avenue for handling diverse types of spectroscopic data, thus expediting the materials characterization processes.journal articl