检索结果-内蒙古大学图书馆

arXiv 2022年

作者： Lee, Ganghun Kim, Minji Lee, Minsu Zhang, Byoung-Tak Interdisciplinary Program in Cognitive Science Seoul National University Seoul Korea Republic of Interdisciplinary Program in Neuroscience Seoul National University Korea Republic of Dept. of Computer Science and Engineering Seoul National University Korea Republic of AIIS Seoul National University Korea Republic of

We present an automated learning framework for a robotic sketching agent that is capable of learning stroke-based rendering and motor control simultaneously. We formulate the robotic sketching problem as a deep decoupled hierarchical reinforcement learning;two policies for stroke-based rendering and motor control are learned independently to achieve sub-tasks for drawing, and form a hierarchy when cooperating for real-world drawing. Without hand-crafted features, drawing sequences or trajectories, and inverse kinematics, the proposed method trains the robotic sketching agent from scratch. We performed experiments with a 6-DoF robot arm with 2F gripper to sketch doodles. Our experimental results show that the two policies successfully learned the sub-tasks and collaborated to sketch the target images. Also, the robustness and flexibility were examined by varying drawing tools and surfaces. Copyright © 2022, The Authors. All rights reserved.

关键词： Robotics

来源：评论

学校读者我要写书评

暂无评论

The Community Human Development Index (CHDI) as a Precision Public Health Vulnerability Metric and Risk Indicator for Predictive Analytics

The Community Human Development Index (CHDI) as a Precision ...

引用

Conference on Technologies for Sustainability (Sustech)

作者： Suraj Sheth Luís Bettencourt Pritzker School of Medicine and Interdisciplinary Scientist Training Program Biomedical Informatics Precision Medicine and Global Health Mansueto Institute for Urban Innovation University of Chicago Chicago IL USA Mansueto Institute for Urban Innovation Ecology and Evolution Dept. of Sociology University of Chicago Chicago IL USA

Accurately assessing and measuring population vulnerability across geographic scales is critical for health risk management of sustainable communities. Currently, the most widely used metric for vulnerability analysis in the United States is the Centers for Disease Control and Prevention's Social Vulnerability Index (SVI), a metric which has proven its utility in a number of circumstances. However, at times, its structure limits its use in international contexts and longitudinal studies. We propose the use of the Community Human Development Index (CHDI), a scalable and predictive tool, to address these gaps. Based on the setting, the CHDI can work either as a complementary metric alongside the SVI or as an alternative vulnerability metric. In our comprehensive study of over 70,000 census tracts in the United States, we found that CHDI correlates well with SVI at the neighborhood scale, while also capturing uniquely different information about the vulnerability of populations. We also found that the CHDI predicts community health outcomes as well or better than the SVI, and can be used as a proxy to measure changes in overall health risks to communities. As a Precision Public Health metric for population health vulnerability analysis and as a performance indicator of community health, CHDI can be used to provide an analytical basis for policy decision making. CHDI can be utilized to promote health equity across a broad range of communities through more accurate health risk assessments leading to strategic, targeted, and timely interventions for better health outcomes.

关键词： Measurement Sociology Decision making Indexes Risk management Statistics Sustainable development

来源：评论

学校读者我要写书评

暂无评论

A Deep Learning Model of Amyloid-${\boldsymbol{\beta }}$ Diffusion Simulation

A Deep Learning Model of Amyloid-${\boldsymbol{\beta }}$ Dif...

引用

International Conference on Electronics, Information and Communications (ICEIC)

作者： ByeongChang Jeong Daegyeom Kim Hyun-Ghang Jeong Cheol E. Han Dept. of Electronic and Information Engineering The Alzheimer's Disease Neuroimaging Initiative Korea University Sejong Republic of Korea The Alzheimer's Disease Neuroimaging Initiative Interdisciplinary Graduate Program for Artificial Intelligence Smart Convergence Technology Korea University Sejong Republic of Korea Dept. of Psychiatry The Alzheimer's Disease Neuroimaging Initiative Korea University College of Medicine Seoul Republic of Korea

Alzheimer's disease (AD) is a severe neurodegenerative disease characterized by ongoing brain tissue decline. Since amyloid- $\beta$ accumulation is supposed as a primary cause of AD progression, mathematical models have been recently employed to simulate amyloid- $\beta$ diffusion in our brain. These models require parameter estimation through observed data and an iterative process to update parameters, although increased complexity can hinder their optimization. To overcome these difficulties, we converted the mathematical model of amyloid- $\beta$ diffusion into a deep learning model, combining multi-layer perceptron (MLP) and graph convolutional neural network (GCN). The proposed model well predicted the change of accumulation level of amyloid- $\beta$ with a high accuracy.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A Study on Improvement of Tomato Disease Classification Performance According to Various Image Augmentation

引用

Transactions of the Korean Institute of Electrical Engineers 2021年第12期70卷 2000-2005页

作者： Ham, Hyun-Sik Cho, Hyun-Chong Dept. of Electronic Engineering Interdisciplinary Graduate Program for BIT Medical Convergence Kangwon National University Korea Republic of Interdisciplinary Graduate Program for BIT Medical Convergence Kangwon National University Korea Republic of

Crop diseases are damaging agricultural food worldwide. Early detection of crop diseases is important to prevent damage to crops. However, it is difficult to distinguish crop diseases unless not expert. Therefore, in this paper, a crop diseases classification system that can recognize diseases even before expert judgment is proposed. The characteristics of diseases were learned and classified using Convolutional Neural Network(CNN). In addition, images are augmented to increase classification performance. To augment the image, a basic augmentation policy consisting of Random Crop and Random Horizontal Flip and Google's AutoAugment are used. Then, the augmented image was selected as an base augmentation model by setting a threshold and then trained. We compared the performance of the original, augmentation model, and image selection model. As a result, the image selection model set to the threshold value of 0.7 in AutoAugment achieved the performance of F1 Score 0.958. © The Korean Institute of Electrical Engineers

关键词： Crops

来源：评论

学校读者我要写书评

暂无评论

A New Image Augmentation Method for Improving Computer-aided Diagnosis System Performance

引用

Transactions of the Korean Institute of Electrical Engineers 2021年第1期70卷 102-107页

作者： Lee, Sin-Ae Cho, Hyun-Chong Interdisciplinary Graduate Program for Bit Medical Convergence Kangwon National University Korea Republic of Dept. of Electronics Engineering and Interdisciplinary Graduate Program for Bit Medical Convergence Kangwon National University Korea Republic of

Gastric cancer is the largest percentage of cancer cases in Korea. A precise way to find the occurrence of gastrointestinal diseases is through gastroscopy by a trained diagnostic physician. Computer-aided diagnosis (CADx) system helps improve the reliability and speed of diagnosis. Hie CADx system has developed with deep learning, which is data dependent. However, medical image data is labor intensive and time consuming, making it difficult for large data sets to be formed. To solve this problem, it is important to apply augmentation techniques. In tliis paper, we propose an augmentation method which is suitable for the data. A basic classification model was made by leaning a data set consisting of the original images. Each of the 14 augment technique-applied data set was input into the generated model and the fl-score values were compared. Hie fl-score of the highest performance among the proposed methods, was 0.9221, with an increase of about 0.085. © 2021 Korean Institute of Electrical Engineers. All rights reserved.

关键词： Computer aided diagnosis

来源：评论

学校读者我要写书评

暂无评论

DCGAN-based Medical Image Augmentation to Improve ELM Classification Performance

DCGAN-based Medical Image Augmentation to Improve ELM Classi...

引用

IEEE International Conference on Communication, Networks and Satellite (ComNetSat)

作者： Rando Noor Akhmad Setiawan Adhistya Erna Permanasari Riries Rulaningtyas Andriyan Bayu Suksmono Imas Sukaesih Sitanggang Dept. of Electrical and Information Engineering Universitas Gadjah Mada Yogyakarta Indonesia Dept. of Physics Universitas Airlangga Biomedical Engineering Program Surabaya Indonesia School of Electrical Engineering and Informatics Institut Teknologi Bandung Bandung Indonesia Department of Computer Science IPB University Bogor Indonesia

ISBN: (数字)9781665460309

ISBN: (纸本)9781665460316

Cervical cancer is one of the deadliest diseases in women. One of the cervical cancer screening methods is pap smear method. However, using a pap smear method to detect cervical cancer takes a long time for a pathologist to diagnose. Hence, a rapid development of medical computerization for early detection to get the results quickly is needed. This paper proposes synthetic data augmentation by using Deep Convolutional Generative Adversarial Network (DCGAN) to increase number of pap smear samples in dataset. Gray Level Co-occurrence Matrix (GLCM) is employed to extract features from dataset. Classification of 3 classes which are Adenocarcinoma, High-Grade Squamous Intraepithelial Lesion (HSIL), and Squamous Cell Carcinoma (SCC) is conducted using Extreme Learning Machine (ELM). The result shows that the addition of synthetic data improves the performance of ELM with the accuracy of 90%. This accuracy is better than the accuracy of ELM using only the original dataset which is 85%.

关键词： Satellites Extreme learning machines Medical services Feature extraction Generative adversarial networks Entropy Lesions

来源：评论

学校读者我要写书评

暂无评论

On the Importance of Critical Period in Multi-stage Reinforcement Learning

arXiv

引用

arXiv 2022年

作者： Park, Junseok Hwang, Inwoo Lee, Min Whoo Oh, Hyunseok Lee, Minsu Lee, Youngki Zhang, Byoung-Tak Interdisciplinary Program In Neuroscience Seoul National University Seoul Korea Republic of Dept. of Computer Science and Engineering Seoul National University Seoul Korea Republic of AIIS Seoul Korea Republic of

The initial years of an infant’s life are known as the critical period, during which the overall development of learning performance is significantly impacted due to neural plasticity. In recent studies, an AI agent, with a deep neural network mimicking mechanisms of actual neurons, exhibited a learning period similar to human’s critical period. Especially during this initial period, the appropriate stimuli play a vital role in developing learning ability. However, transforming human cognitive bias into an appropriate shaping reward is quite challenging, and prior works on critical period do not focus on finding the appropriate stimulus. To take a step further, we propose multi-stage reinforcement learning to emphasize finding "appropriate stimulus" around the critical period. Inspired by humans’ early cognitive-developmental stage, we use multi-stage guidance near the critical period, and demonstrate the appropriate shaping reward (stage-2 guidance) in terms of the AI agent’s performance, efficiency, and stability. © 2022, CC BY-NC-SA.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Combining Knee Acoustic Emissions, Patient-Reported Measures, and Machine Learning to Assess Osteoarthritis Severity

Combining Knee Acoustic Emissions, Patient-Reported Measures...

引用

International Workshop on Wearable and Implantable Body Sensor Networks (BSN)

作者： Christopher Nichols H. Trask Crane Dave Ewart Omer T. Inan Bioengineering Graduate Program Georgia Institute of Technology Atlanta GA USA School of Electrical and Computer Engineering Georgia Institute of Technology Atlanta GA USA Minneapolis Veterans Affairs Medical Center Minneapolis MN USA Coulter Dept. of Biomedical Engineering School of Electrical and Computer Engineering Georgia Institute of Technology Atlanta GA USA

Current methods for quantifying osteoarthritis severity have limited resolution and accessibility. Patient-recorded outcome measures such as the Knee Injury and Osteoarthritis Outcome Score (KOOS) capture symptom severity, but are subjectively reported and have little correlation with quantifiable metrics of disease such as Kellgren-Lawrence x-ray grade or MRI findings. Knee acoustic emissions (KAEs) offer a convenient, noninvasive option for quantifying joint health. Here, we use machine learning and wearable design to create an interpretable two-stage algorithm for combining KAEs and KOOS scores into an objective, more accessible method of quantifying disease severity. Our algorithm successfully discriminated between early and late-stage osteoarthritis (balanced accuracy = 85%, ROC-AUC = 0.88). The addition of KAEs improved classification of osteoarthritis severity over the use of KAEs (balanced accuracy = 53%, ROC-AUC = 0.786) or KOOS scores alone (balanced accuracy = 63%, ROC-AUC = 0.593). The findings suggest that KAEs combined with patient-recorded metrics can be used to make a more objective and accessible metric for digitally monitoring knee joint health.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Accelerating Real-Time Imaging for Radiotherapy: Leveraging Multi-GPU Training with PyTorch

Accelerating Real-Time Imaging for Radiotherapy: Leveraging ...

引用

International Conference on Machine Learning and Applications (ICMLA)

作者： Ruth Obe Brandt Kaufmann Kaelen Baird Sam Kadel Yasmin Soltani Mostafa Cham Matthias K. Gobbert Carlos A. Barajas Zhuoran Jiang Vijay R. Sharma Lei Ren Stephen W. Peterson Jerimy C. PoIf Dept. of Computer Science and of Software Engineering U. of Houston-Clear Lake USA Dept. of Mathematics and Statistics U. of San Francisco USA Dept. of Computer Science and of Mathematics Skidmore College USA Dept. of Computer Science and of Psychology Mount Holyoke College USA Dept. of Biomedical Engineering U. of Houston USA Dept. of Information Systems U. of Maryland Baltimore County Dept. of Mathematics and Statistics U. of Maryland Baltimore County Medical Physics Graduate Program Duke University USA Dept. of Radiation Oncology U. of Maryland School of Medicine USA Department of Radiation Oncology U. of Maryland School of Medicine USA Dept. of Physics U. of Cape Town South Africa H3D Inc. USA

Proton beam therapy is an advanced form of cancer radiotherapy that uses high-energy proton beams to deliver precise and targeted radiation to tumors. This helps to mit-igate unnecessary radiation exposure in healthy tissues. Real-time imaging of prompt gamma rays with Compton cameras has been suggested to improve therapy efficacy. However, the camera's non-zero time resolution leads to incorrect interaction classifications and noisy images that are insufficient for accurately assessing proton delivery in patients. To address the challenges posed by the Compton camera's image quality, machine learning techniques are employed to classify and refine the generated data. These machine-learning techniques include recurrent and feedforward neural networks. A PyTorch model was designed to improve the data captured by the Compton camera. This decision was driven by PyTorch's flexibility, powerful capabilities in handling sequential data, and enhanced G PU usage. This accelerates the model's computations on large-scale radiotherapy data. Through hyperparameter tuning, the validation accuracy of our PyTorch model has been improved from an initial 7 % to over 60 %. Moreover, the PyTorch Distributed Data Parallelism strategy was used to train the RNN models on multiple G PU s, which significantly reduced the training time with a minor impact on model accuracy.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Mammographic Breast Positioning Assessment via Deep Learning

arXiv

引用

arXiv 2024年

作者： Tanyel, Toygar Denizoglu, Nurper Seker, Mustafa Ege Alis, Deniz Cerekci, Esma Karaarslan, Ercan Aribal, Erkin Oksuz, Ilkay Istanbul Technical University Biomedical Engineering Graduate Program Turkey Acibadem Healthcare Group Department of Radiology Turkey Acibadem Mehmet Ali Aydinlar University School of Medicine Turkey Acibadem Mehmet Ali Aydinlar University School of Medicine Dept. of Radiology Turkey Sisli Hamidiye Etfal Training and Research Hospital Turkey Istanbul Technical University Department of Computer Engineering Turkey

Breast cancer remains a leading cause of cancer-related deaths among women worldwide, with mammography screening as the most effective method for the early detection. Ensuring proper positioning in mammography is critical, as poor positioning can lead to diagnostic errors, increased patient stress, and higher costs due to recalls. Despite advancements in deep learning (DL) for breast cancer diagnostics, limited focus has been given to evaluating mammography positioning. This paper introduces a novel DL methodology to quantitatively assess mammogram positioning quality, specifically in mediolateral oblique (MLO) views using attention and coordinate convolution modules. Our method identifies key anatomical landmarks, such as the nipple and pectoralis muscle, and automatically draws a posterior nipple line (PNL), offering robust and inherently explainable alternative to well-known classification and regression-based approaches. We compare the performance of proposed methodology with various regression and classification-based models. The CoordAtt UNet model achieved the highest accuracy of 88.63% ± 2.84 and specificity of 90.25% ± 4.04, along with a noteworthy sensitivity of 86.04% ± 3.41. In landmark detection, the same model also recorded the lowest mean errors in key anatomical points and the smallest angular error of 2.42 degrees. Our results indicate that models incorporating attention mechanisms and CoordConv module increase the accuracy in classifying breast positioning quality and detecting anatomical landmarks. Furthermore, we make the labels and source codes available to the community to initiate an open research area for mammography, accessible at https://***/tanyelai/deep-breast-positioning. © 2024, CC BY-NC-SA.

关键词： Diseases

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：