Ngọc Định Nguyễn - Academia.edu (original) (raw)

Papers by Ngọc Định Nguyễn

Research paper thumbnail of Efficient Feedback and Partial Credit Grading for Proof Blocks Problems

Lecture Notes in Computer Science

Research paper thumbnail of Bất Thường Nhiễm Sắc Thể Phôi Theo Nhóm Tuổi Của Bệnh Nhân Vô Sinh Điều Trị Thụ Tinh Trong Ống Nghiệm

Tạp chí Y học Việt Nam, Apr 20, 2023

Mục tiêu: Đánh giá bất thường nhiễm sắc thể (NST) phôi theo nhóm tuổi của bệnh nhân vô sinh có ch... more Mục tiêu: Đánh giá bất thường nhiễm sắc thể (NST) phôi theo nhóm tuổi của bệnh nhân vô sinh có chỉ định điều trị thụ tinh trong ống nghiệm kết hợp sàng lọc di truyền trước chuyển phôi. Đối tượng và phương pháp nghiên cứu: Mô tả tiến cứu trên 661 phôi túi của 186 bệnh nhân hiếm muộn có chỉ định điều trị thụ tinh trong ống nghiệm (In vitro fertilization/IVF) kết hợp sàng lọc di truyền trước chuyển phôi ((Preimplantation genetic testing for aneuploidies/PGT-A). Kết quả: Tỉ lệ phôi mang NST bình thường tương đương nhau giữa nhóm I (dưới 30 tuổi) và nhóm II (30-35 tuổi) lần lượt là 51,1 và 54,1% (P(1-2)=0,613); thấp nhất ở nhóm III (>35 tuổi) với tỉ lệ 35,7%, P(1-3); P(2-3) <0,05. Tỉ lệ phôi bất thường NST khác biệt không có ý nghĩa giữa nhóm I và nhóm II (P(1-2)=0,310), tăng cao có ý nghĩa thống kê ở nhóm III với tỉ lệ là 47,2% (P(1-3); P(2-3)<0,001). Bất thường số lượng NST có xu hướng tăng theo sự gia tăng của tuổi mẹ (p<0,05), trong khi đó bất thường cấu trúc NST xu hướng ngược lại mặc dù chưa thấy sự khác biệt có ý nghĩa (p=0,148). Tỉ lệ phôi khảm ở ba nhóm lần lượt là 25%;16,5%; 17,1% (P(1-2)=0,069; P(1-3)=0,101; P(2-3)=0,843). Trong các phôi bất thường số lượng NST, tỉ lệ phôi monosomy và trisomy là tương đương ở ba nhóm (P(1-2);P(1-3); P(2-3)>0,05). Kết luận: Trên đối tượng bệnh nhân vô sinh có chỉ định điều trị IVF-PGT-A, tỉ lệ phôi bất thường NST tăng có ý nghĩa thống kê khi tuổi người mẹ trên

Research paper thumbnail of Neural Inverse Text Normalization with Numerical Recognition for Low Resource Scenarios

Lecture Notes in Computer Science, 2022

Research paper thumbnail of Giá Trị Của Phương Pháp Chọc Hút Dưới Hướng Dẫn Nội Soi Phế Quản Siêu Âm Trong Chẩn Đoán Nguyên Nhân Các Tổn Thương Của Trung Thất, Rốn Phổi

Tạp chí Y học Việt Nam, Feb 15, 2023

Chọc hút tổn thương dưới hướng dẫn nội soi phế quản siêu âm (EBUS-TBNA) dùng để chẩn đoán các tổn... more Chọc hút tổn thương dưới hướng dẫn nội soi phế quản siêu âm (EBUS-TBNA) dùng để chẩn đoán các tổn thương trung thất, rốn phổi. Chúng tôi nghiên cứu giá trị chẩn đoán và độ an toàn của thủ thuật như là một phương pháp chẩn đoán ban đầu. Phương pháp nghiên cứu: nghiên cứu hồi cứu mô tả cắt ngang từ tháng 9/2017 tới tháng 7/2021. Kết quả: EBUS-TBNA có độ chẩn đoán chính xác 69%; độ nhạy, đặc hiệu, giá trị dự báo dương tính, giá trị dự báo âm tính lần lượt là: 62%, 100%, 100%, và 38.9%. Không có biến chứng nào được ghi nhận. Kết luận: EBUS-TBNA là thủ thuật an toàn, hiệu quả, nên được sử dụng như một dụng cụ ban đầu để chẩn đoán các tổn thương trung thất, rốn phổi. Từ khoá: Chọc hút dưới hướng dẫn nội soi phế quản siêu âm (EBUS-TBNA), tổn thương trung thất, rốn phổi. SUMMARY DIAGNOSTIC VALUE OF ENDOBRONCHIAL ULTRASOUND-GUIDED TRANSBRONCHIAL NEEDLE ASPIRATION (EBUS-TBNA) IN DIAGNOSING ENLARGED MEDIASTINAL/HILAR LESIONS Real-time, convex probe endobronchial ultrasoundguided transbronchial needle aspiration (EBUS-TBNA) is used to reveal etiologies of enlarged mediastinal/hilar lymphadenopathy, lesions. The study evaluated the diagnostic efficacy and safety of EBUS-TBNA when used as an initial diagnostic tool. We restrospective 75 patients between September 2017 and July 2021. Results: In the 72 cases with adequate results, the diagnostic accuracy of EBUS-BNA was calculated 69%; the diagnostics specific, the sensitive, the positive predictive value, the negative predictive value was 62%, 100%, 38.9%, and 100%. There was no reported complication. Conclusions: EBUS-TBNA is safe and efficacy procedure, so EBUS-TBNA should be used as an initial diagnostic tool for the assessment of mediastinal and hilar lymph nodes, lesions.

Research paper thumbnail of Giá Trị Của Cộng Hưởng Từ Động Sàn Chậu Trong Chẩn Đoán Hội Chứng Đại Tiện Tắc Nghẽn

Tạp chí Điện quang Việt Nam, Nov 30, 2022

Research paper thumbnail of Towards designing a generic and comprehensive deep reinforcement learning framework

Applied Intelligence

Reinforcement learning (RL) has emerged as an effective approach for building an intelligent syst... more Reinforcement learning (RL) has emerged as an effective approach for building an intelligent system, which involves multiple self-operated agents to collectively accomplish a designated task. More importantly, there has been a renewed focus on RL since the introduction of deep learning that essentially makes RL feasible to operate in high-dimensional environments. However, there are many diversified research directions in the current literature, such as multi-agent and multi-objective learning, and human-machine interactions. Therefore, in this paper, we propose a comprehensive software architecture that not only plays a vital role in designing a connect-the-dots deep RL architecture but also provides a guideline to develop a realistic RL application in a short time span. By inheriting the proposed architecture, software managers can foresee any challenges when designing a deep RL-based system. As a result, they can expedite the design process and actively control every stage of sof...

Research paper thumbnail of Grain-boundary segregation of magnesium in doped cuprous oxide and impact on electrical transport properties

Scientific Reports

In this study, we report the segregation of magnesium in the grain boundaries of magnesium-doped ... more In this study, we report the segregation of magnesium in the grain boundaries of magnesium-doped cuprous oxide (Cu2O:Mg) thin films as revealed by atom probe tomography and the consequences of the dopant presence on the temperature-dependent Hall effect properties. The incorporation of magnesium as a divalent cation was achieved by aerosol-assisted metal organic chemical vapour deposition, followed by thermal treatments under oxidizing conditions. We observe that, in comparison with intrinsic cuprous oxide, the electronic transport is improved in Cu2O:Mg with a reduction of resistivity to 13.3 ± 0.1 Ω cm, despite the reduction of hole mobility in the doped films, due to higher grain-boundary scattering. The Hall carrier concentration dependence with temperature showed the presence of an acceptor level associated with an ionization energy of 125 ± 9 meV, similar to the energy value of a large size impurity−vacancy complex. Atom probe tomography shows a magnesium incorporation of 5%, ...

Research paper thumbnail of Fruit-CoV: An efficient vision-based framework for speedy detection and diagnosis of SARS-CoV-2 infections through recorded cough sounds

Expert Systems with Applications

SARS-CoV-2 is colloquially known as COVID-19 that had an initial outbreak in December 2019. The d... more SARS-CoV-2 is colloquially known as COVID-19 that had an initial outbreak in December 2019. The deadly virus has spread across the world, taking part in the global pandemic disease since March 2020. In addition, a recent variant of SARS-CoV-2 named Delta is intractably contagious and responsible for more than four million deaths over the world. Therefore, it is vital to possess a self-testing service of SARS-CoV-2 at home. In this study, we introduce Fruit-CoV, a two-stage vision framework, which is capable of detecting SARS-CoV-2 infections through recorded cough sounds. Specifically, we convert sounds into Log-Mel Spectrograms and use the EfficientNet-V2 network to extract its visual features in the first stage. In the second stage, we use 14 convolutional layers extracted from the large-scale Pretrained Audio Neural Networks for audio pattern recognition (PANNs) and the Wavegram-Log-Mel-CNN to aggregate feature representations of the Log-Mel Spectrograms. Finally, we use the combined features to train a binary classifier. In this study, we use a dataset provided by the AICovidVN 115M Challenge, which includes a total of 7371 recorded cough sounds collected throughout Vietnam, India, and Switzerland. Experimental results show that our proposed model achieves an AUC score of 92.8% and ranks the 1st place on the leaderboard of the AICovidVN Challenge. More importantly, our proposed framework can be integrated into a call center or a VoIP system to speed up detecting SARS-CoV-2 infections through online/recorded cough sounds.

Research paper thumbnail of Nghiên cứu thành lập chương trình xử lý số liệu quan trắc và dự báo độ lún công trình

Tạp chí Khoa học Đo đạc và Bản đồ, 2016

Tính toán và xử lý số liệu quan trắc lún công trình là một công tác đòi hỏi yêu cầu độ chính xác ... more Tính toán và xử lý số liệu quan trắc lún công trình là một công tác đòi hỏi yêu cầu độ chính xác và cung cấp số liệu nhanh chóng trong quá trình thi công và sử dụng công trình. Bởi vậy, việc nghiên cứu thành lập chương trình xử lý số liệu quan trắc và dự báo độ lún công trình đáp ứng các yêu cầy trên là việc làm cần thiết, có ý nghĩa khoa học và thực tiễn cao. Bài báo giới thiệu kết quả nghiên cứu của nhóm tác giả thực hiện đề tài về xây dựng phần mềm Eng-Survey nhằm giải quyết một số vấn đề của công tác trắc địa trong đó có tính năng xử lý số liệu đo lún công trình.

Research paper thumbnail of Author Correction: Grain-boundary segregation of magnesium in doped cuprous oxide and impact on electrical transport properties

Research paper thumbnail of Examination of the Antecedents, Reactions, and Outcomes to a Major Technology-driven Organizational Change

The goal of this study was to test a multi-level model of organizational change that examined how... more The goal of this study was to test a multi-level model of organizational change that examined how various antecedents, employee reactions, and organizational and personal outcomes relate to one another. The research was conducted via online surveys and as a longitudinal study. Participants were employees at a large supply distribution company, and were a part of the Pilot implementation of a new Enterprise Resource Planning (ERP) system. Results from the study revealed that job stress was closely related to organizational commitment, job satisfaction, and psychological well-being, while change commitment was associated with higher organizational commitment and job satisfaction. Positive training reactions were linked to increased change commitment and organizational commitment, and change-specific self-efficacy also predicted commitment to change. Additionally, change self-efficacy and principal support significantly moderated the relationship between coping and organizational commi...

Research paper thumbnail of Trapping of Electrons around Nanoscale Metallic Wires Embedded in a Semiconductor Medium

arXiv: Mesoscale and Nanoscale Physics, 2019

We predict that conduction electrons in a semiconductor film containing a centered square array o... more We predict that conduction electrons in a semiconductor film containing a centered square array of metal nanowires normal to its plane are bound in quantum states around the central wires, if a positive bias voltage is applied between the wires at the square vertices and these latter. We obtain and discuss the eigenenergies and eigenfunctions of two models with different dimensions. The results show that the eigenstates can be grouped into different shells. The energy differences between the shells is typically a few tens of meV, which corresponds to frequencies of emitted or absorbed photons in a range of 3 THz to 20 THz approximately. These energy differences strongly depend on the bias voltage. We calculate the linear response of individual electrons on the ground level of our models to large-wavelength electromagnetic waves whose electric field is in the plane of the semiconductor film. The computed oscillator strengths are dominated by the transitions to the states in each shel...

Research paper thumbnail of A simple printed antenna with broadband property and omnidirectional radiation patterns of wire dipole

IEICE Electron. Express, 2020

This letter presents a unique printed dipole antenna which can roughly cover an 80% fractional ba... more This letter presents a unique printed dipole antenna which can roughly cover an 80% fractional bandwidth (VSWR  3), and especially possess the omnidirectional radiation patterns similar to those of wire dipole antenna at its all resonant modes. The antenna is simple with a microstrip-fed structure which it is just composed of two arms with each one forming the shape of a conventional "inset-fed" rectangular patch antenna. At the lowest operation frequency for VSWR = 3, the antenna size is only 0.3λ × 0.008λ, showing its effectiveness in many applications of small antennas.

Research paper thumbnail of Review, Analyze, and Design a Comprehensive Deep Reinforcement Learning Framework

ArXiv, 2020

Reinforcement learning (RL) has emerged as a standard approach for building an intelligent system... more Reinforcement learning (RL) has emerged as a standard approach for building an intelligent system, which involves multiple self-operated agents to collectively accomplish a designated task. More importantly, there has been a great attention to RL since the introduction of deep learning that essentially makes RL feasible to operate in high-dimensional environments. However, current research interests are diverted into different directions, such as multi-agent and multi-objective learning, and human-machine interactions. Therefore, in this paper, we propose a comprehensive software architecture that not only plays a vital role in designing a connect-the-dots deep RL architecture but also provides a guideline to develop a realistic RL application in a short time span. By inheriting the proposed architecture, software managers can foresee any challenges when designing a deep RL-based system. As a result, they can expedite the design process and actively control every stage of software d...

Research paper thumbnail of A Visual Communication Map for Multi-Agent Deep Reinforcement Learning

ArXiv, 2020

Multi-agent learning distinctly poses significant challenges in the effort to allocate a conceale... more Multi-agent learning distinctly poses significant challenges in the effort to allocate a concealed communication medium. Agents receive thorough knowledge from the medium to determine subsequent actions in a distributed nature. Apparently, the goal is to leverage the cooperation of multiple agents to achieve a designated objective efficiently. Recent studies typically combine a specialized neural network with reinforcement learning to enable communication between agents. This approach, however, limits the number of agents or necessitates the homogeneity of the system. In this paper, we have proposed a more scalable approach that not only deals with a great number of agents but also enables collaboration between dissimilar functional agents and compatibly combined with any deep reinforcement learning methods. Specifically, we create a global communication map to represent the status of each agent in the system visually. The visual map and the environmental state are fed to a shared-p...

Research paper thumbnail of A Prioritized objective actor-critic method for deep reinforcement learning

Neural Computing and Applications, 2021

An increasing number of complex problems have naturally posed significant challenges in decision-... more An increasing number of complex problems have naturally posed significant challenges in decision-making theory and reinforcement learning practices. These problems often involve multiple conflicting reward signals that inherently cause agents’ poor exploration in seeking a specific goal. In extreme cases, the agent gets stuck in a sub-optimal solution and starts behaving harmfully. To overcome such obstacles, we introduce two actor-critic deep reinforcement learning methods, namely Multi-Critic Single Policy (MCSP) and Single Critic Multi-Policy (SCMP), which can adjust agent behaviors to efficiently achieve a designated goal by adopting a weighted-sum scalarization of different objective functions. In particular, MCSP creates a human-centric policy that corresponds to a predefined priority weight of different objectives. Whereas, SCMP is capable of generating a mixed policy based on a set of priority weights, i.e., the generated policy uses the knowledge of different policies (each policy corresponds to a priority weight) to dynamically prioritize objectives in real time. We examine our methods by using the Asynchronous Advantage Actor-Critic (A3C) algorithm to utilize the multithreading mechanism for dynamically balancing training intensity of different policies into a single network. Finally, simulation results show that MCSP and SCMP significantly outperform A3C with respect to the mean of total rewards in two complex problems: Food Collector and Seaquest.

Research paper thumbnail of Manipulating Soft Tissues by Deep Reinforcement Learning for Autonomous Robotic Surgery

2019 IEEE International Systems Conference (SysCon), 2019

In robotic surgery, pattern cutting through a deformable material is a challenging research field... more In robotic surgery, pattern cutting through a deformable material is a challenging research field. The cutting procedure requires a robot to concurrently manipulate a scissor and a gripper to cut through a predefined contour trajectory on the deformable sheet. The gripper ensures the cutting accuracy by nailing a point on the sheet and continuously tensioning the pinch point to different directions while the scissor is in action. The goal is to find a pinch point and a corresponding tensioning policy to minimize damage to the material and increase cutting accuracy measured by the symmetric difference between the predefined contour and the cut contour. Previous study considers finding one fixed pinch point during the course of cutting, which is inaccurate and unsafe when the contour trajectory is complex. In this paper, we examine the soft tissue cutting task by using multiple pinch points, which imitates human operations while cutting. This approach, however, does not require the use of a multi-gripper robot. We use a deep reinforcement learning algorithm to find an optimal tensioning policy of a pinch point. Simulation results show that the multi-point approach outperforms the state-ofthe-art method in soft pattern cutting task with respect to both accuracy and reliability.

Research paper thumbnail of Application of St Gallen Categories in Predicting Survival for Patients With Breast Cancer in Vietnam

Cancer Control, 2019

Breast cancer is a heterogeneous disease with different tumor subtypes. Identifying risk categori... more Breast cancer is a heterogeneous disease with different tumor subtypes. Identifying risk categories will help make better treatment decisions. Hence, this study aimed to predict the survival outcomes of invasive breast cancer in Vietnam, using St Gallen 2007 classification. This study was conducted on 501 patients with breast cancer who had surgical operations, but had not received neoadjuvant chemotherapy, from 2011 to 2013. The clinicopathological characteristics were recorded. Immunohistochemistry staining was performed on ER, PR, HER2/neu, and Ki67 markers. For HER2/neu(2+), fluorescence in situ hybridization was used as the test. All patients with breast cancer were stratified according to 2007 St Gallen categories. Kaplan-Meier and log-rank models were used to analyze survival rates. There were 3.8% cases classified as low risk (LR), 72.1% as intermediate risk (IR1: 60.1% and IR2: 12.0%), and 24.1% as high risk (HR1: 11.8% and HR2: 12.3%). Patients who were LR had the best pro...

Research paper thumbnail of System Design Perspective for Human-Level Agents Using Deep Reinforcement Learning: A Survey

IEEE Access, 2017

Reinforcement learning (RL) has distinguished itself as a prominent learning method to augment th... more Reinforcement learning (RL) has distinguished itself as a prominent learning method to augment the efficacy of autonomous systems. Recent advances in deep learning studies have complemented existing RL methods and led to a crucial breakthrough in the effort of applying RL to automation and robotics. Artificial agents based on deep RL can take selective and intelligent actions comparable with those of a human to maximize the feedback reward from the interactive environment. In this paper, we survey recent developments in the literature regarding deep RL methods for building human-level agents. As a result, prominent studies that involve modeling every aspect of a human-level agent will be examined. We also provide an overview of constructing a framework for prospective autonomous systems. Moreover, various toolkits and frameworks are suggested to facilitate the development of deep RL methods. Finally, we open a discussion that potentially raises a range of future research directions in deep RL. INDEX TERMS Deep learning, human-level agents, reinforcement learning, robotics, survey, system design.

Research paper thumbnail of Investor confidence and mutual fund performance in emerging markets

Journal of Economic Studies, 2018

Purpose The purpose of this paper is to investigate the impact of investor confidence on mutual f... more Purpose The purpose of this paper is to investigate the impact of investor confidence on mutual fund performance in two relatively vulnerable but leading emerging markets, India and Pakistan. Design/methodology/approach A pooled ordinary least squared (OLS) model is used to look at two alternative measures of investor confidence and test for the relationship between investor confidence and mutual fund returns. To check the robustness of the findings, the authors also implement two-stage least squares and generalized method of moments techniques to control for unobserved heterogeneity, simultaneity and dynamic endogeneity problems in the regressors. Findings The paper finds that the returns of mutual funds are positively associated with investor confidence and an interaction effect exists between investor confidence and persistence in performance. The paper also confirms that returns from mutual funds are associated with different fund characteristics such as fund size, turnover, exp...

Research paper thumbnail of Efficient Feedback and Partial Credit Grading for Proof Blocks Problems

Lecture Notes in Computer Science

Research paper thumbnail of Bất Thường Nhiễm Sắc Thể Phôi Theo Nhóm Tuổi Của Bệnh Nhân Vô Sinh Điều Trị Thụ Tinh Trong Ống Nghiệm

Tạp chí Y học Việt Nam, Apr 20, 2023

Mục tiêu: Đánh giá bất thường nhiễm sắc thể (NST) phôi theo nhóm tuổi của bệnh nhân vô sinh có ch... more Mục tiêu: Đánh giá bất thường nhiễm sắc thể (NST) phôi theo nhóm tuổi của bệnh nhân vô sinh có chỉ định điều trị thụ tinh trong ống nghiệm kết hợp sàng lọc di truyền trước chuyển phôi. Đối tượng và phương pháp nghiên cứu: Mô tả tiến cứu trên 661 phôi túi của 186 bệnh nhân hiếm muộn có chỉ định điều trị thụ tinh trong ống nghiệm (In vitro fertilization/IVF) kết hợp sàng lọc di truyền trước chuyển phôi ((Preimplantation genetic testing for aneuploidies/PGT-A). Kết quả: Tỉ lệ phôi mang NST bình thường tương đương nhau giữa nhóm I (dưới 30 tuổi) và nhóm II (30-35 tuổi) lần lượt là 51,1 và 54,1% (P(1-2)=0,613); thấp nhất ở nhóm III (>35 tuổi) với tỉ lệ 35,7%, P(1-3); P(2-3) <0,05. Tỉ lệ phôi bất thường NST khác biệt không có ý nghĩa giữa nhóm I và nhóm II (P(1-2)=0,310), tăng cao có ý nghĩa thống kê ở nhóm III với tỉ lệ là 47,2% (P(1-3); P(2-3)<0,001). Bất thường số lượng NST có xu hướng tăng theo sự gia tăng của tuổi mẹ (p<0,05), trong khi đó bất thường cấu trúc NST xu hướng ngược lại mặc dù chưa thấy sự khác biệt có ý nghĩa (p=0,148). Tỉ lệ phôi khảm ở ba nhóm lần lượt là 25%;16,5%; 17,1% (P(1-2)=0,069; P(1-3)=0,101; P(2-3)=0,843). Trong các phôi bất thường số lượng NST, tỉ lệ phôi monosomy và trisomy là tương đương ở ba nhóm (P(1-2);P(1-3); P(2-3)>0,05). Kết luận: Trên đối tượng bệnh nhân vô sinh có chỉ định điều trị IVF-PGT-A, tỉ lệ phôi bất thường NST tăng có ý nghĩa thống kê khi tuổi người mẹ trên

Research paper thumbnail of Neural Inverse Text Normalization with Numerical Recognition for Low Resource Scenarios

Lecture Notes in Computer Science, 2022

Research paper thumbnail of Giá Trị Của Phương Pháp Chọc Hút Dưới Hướng Dẫn Nội Soi Phế Quản Siêu Âm Trong Chẩn Đoán Nguyên Nhân Các Tổn Thương Của Trung Thất, Rốn Phổi

Tạp chí Y học Việt Nam, Feb 15, 2023

Chọc hút tổn thương dưới hướng dẫn nội soi phế quản siêu âm (EBUS-TBNA) dùng để chẩn đoán các tổn... more Chọc hút tổn thương dưới hướng dẫn nội soi phế quản siêu âm (EBUS-TBNA) dùng để chẩn đoán các tổn thương trung thất, rốn phổi. Chúng tôi nghiên cứu giá trị chẩn đoán và độ an toàn của thủ thuật như là một phương pháp chẩn đoán ban đầu. Phương pháp nghiên cứu: nghiên cứu hồi cứu mô tả cắt ngang từ tháng 9/2017 tới tháng 7/2021. Kết quả: EBUS-TBNA có độ chẩn đoán chính xác 69%; độ nhạy, đặc hiệu, giá trị dự báo dương tính, giá trị dự báo âm tính lần lượt là: 62%, 100%, 100%, và 38.9%. Không có biến chứng nào được ghi nhận. Kết luận: EBUS-TBNA là thủ thuật an toàn, hiệu quả, nên được sử dụng như một dụng cụ ban đầu để chẩn đoán các tổn thương trung thất, rốn phổi. Từ khoá: Chọc hút dưới hướng dẫn nội soi phế quản siêu âm (EBUS-TBNA), tổn thương trung thất, rốn phổi. SUMMARY DIAGNOSTIC VALUE OF ENDOBRONCHIAL ULTRASOUND-GUIDED TRANSBRONCHIAL NEEDLE ASPIRATION (EBUS-TBNA) IN DIAGNOSING ENLARGED MEDIASTINAL/HILAR LESIONS Real-time, convex probe endobronchial ultrasoundguided transbronchial needle aspiration (EBUS-TBNA) is used to reveal etiologies of enlarged mediastinal/hilar lymphadenopathy, lesions. The study evaluated the diagnostic efficacy and safety of EBUS-TBNA when used as an initial diagnostic tool. We restrospective 75 patients between September 2017 and July 2021. Results: In the 72 cases with adequate results, the diagnostic accuracy of EBUS-BNA was calculated 69%; the diagnostics specific, the sensitive, the positive predictive value, the negative predictive value was 62%, 100%, 38.9%, and 100%. There was no reported complication. Conclusions: EBUS-TBNA is safe and efficacy procedure, so EBUS-TBNA should be used as an initial diagnostic tool for the assessment of mediastinal and hilar lymph nodes, lesions.

Research paper thumbnail of Giá Trị Của Cộng Hưởng Từ Động Sàn Chậu Trong Chẩn Đoán Hội Chứng Đại Tiện Tắc Nghẽn

Tạp chí Điện quang Việt Nam, Nov 30, 2022

Research paper thumbnail of Towards designing a generic and comprehensive deep reinforcement learning framework

Applied Intelligence

Reinforcement learning (RL) has emerged as an effective approach for building an intelligent syst... more Reinforcement learning (RL) has emerged as an effective approach for building an intelligent system, which involves multiple self-operated agents to collectively accomplish a designated task. More importantly, there has been a renewed focus on RL since the introduction of deep learning that essentially makes RL feasible to operate in high-dimensional environments. However, there are many diversified research directions in the current literature, such as multi-agent and multi-objective learning, and human-machine interactions. Therefore, in this paper, we propose a comprehensive software architecture that not only plays a vital role in designing a connect-the-dots deep RL architecture but also provides a guideline to develop a realistic RL application in a short time span. By inheriting the proposed architecture, software managers can foresee any challenges when designing a deep RL-based system. As a result, they can expedite the design process and actively control every stage of sof...

Research paper thumbnail of Grain-boundary segregation of magnesium in doped cuprous oxide and impact on electrical transport properties

Scientific Reports

In this study, we report the segregation of magnesium in the grain boundaries of magnesium-doped ... more In this study, we report the segregation of magnesium in the grain boundaries of magnesium-doped cuprous oxide (Cu2O:Mg) thin films as revealed by atom probe tomography and the consequences of the dopant presence on the temperature-dependent Hall effect properties. The incorporation of magnesium as a divalent cation was achieved by aerosol-assisted metal organic chemical vapour deposition, followed by thermal treatments under oxidizing conditions. We observe that, in comparison with intrinsic cuprous oxide, the electronic transport is improved in Cu2O:Mg with a reduction of resistivity to 13.3 ± 0.1 Ω cm, despite the reduction of hole mobility in the doped films, due to higher grain-boundary scattering. The Hall carrier concentration dependence with temperature showed the presence of an acceptor level associated with an ionization energy of 125 ± 9 meV, similar to the energy value of a large size impurity−vacancy complex. Atom probe tomography shows a magnesium incorporation of 5%, ...

Research paper thumbnail of Fruit-CoV: An efficient vision-based framework for speedy detection and diagnosis of SARS-CoV-2 infections through recorded cough sounds

Expert Systems with Applications

SARS-CoV-2 is colloquially known as COVID-19 that had an initial outbreak in December 2019. The d... more SARS-CoV-2 is colloquially known as COVID-19 that had an initial outbreak in December 2019. The deadly virus has spread across the world, taking part in the global pandemic disease since March 2020. In addition, a recent variant of SARS-CoV-2 named Delta is intractably contagious and responsible for more than four million deaths over the world. Therefore, it is vital to possess a self-testing service of SARS-CoV-2 at home. In this study, we introduce Fruit-CoV, a two-stage vision framework, which is capable of detecting SARS-CoV-2 infections through recorded cough sounds. Specifically, we convert sounds into Log-Mel Spectrograms and use the EfficientNet-V2 network to extract its visual features in the first stage. In the second stage, we use 14 convolutional layers extracted from the large-scale Pretrained Audio Neural Networks for audio pattern recognition (PANNs) and the Wavegram-Log-Mel-CNN to aggregate feature representations of the Log-Mel Spectrograms. Finally, we use the combined features to train a binary classifier. In this study, we use a dataset provided by the AICovidVN 115M Challenge, which includes a total of 7371 recorded cough sounds collected throughout Vietnam, India, and Switzerland. Experimental results show that our proposed model achieves an AUC score of 92.8% and ranks the 1st place on the leaderboard of the AICovidVN Challenge. More importantly, our proposed framework can be integrated into a call center or a VoIP system to speed up detecting SARS-CoV-2 infections through online/recorded cough sounds.

Research paper thumbnail of Nghiên cứu thành lập chương trình xử lý số liệu quan trắc và dự báo độ lún công trình

Tạp chí Khoa học Đo đạc và Bản đồ, 2016

Tính toán và xử lý số liệu quan trắc lún công trình là một công tác đòi hỏi yêu cầu độ chính xác ... more Tính toán và xử lý số liệu quan trắc lún công trình là một công tác đòi hỏi yêu cầu độ chính xác và cung cấp số liệu nhanh chóng trong quá trình thi công và sử dụng công trình. Bởi vậy, việc nghiên cứu thành lập chương trình xử lý số liệu quan trắc và dự báo độ lún công trình đáp ứng các yêu cầy trên là việc làm cần thiết, có ý nghĩa khoa học và thực tiễn cao. Bài báo giới thiệu kết quả nghiên cứu của nhóm tác giả thực hiện đề tài về xây dựng phần mềm Eng-Survey nhằm giải quyết một số vấn đề của công tác trắc địa trong đó có tính năng xử lý số liệu đo lún công trình.

Research paper thumbnail of Author Correction: Grain-boundary segregation of magnesium in doped cuprous oxide and impact on electrical transport properties

Research paper thumbnail of Examination of the Antecedents, Reactions, and Outcomes to a Major Technology-driven Organizational Change

The goal of this study was to test a multi-level model of organizational change that examined how... more The goal of this study was to test a multi-level model of organizational change that examined how various antecedents, employee reactions, and organizational and personal outcomes relate to one another. The research was conducted via online surveys and as a longitudinal study. Participants were employees at a large supply distribution company, and were a part of the Pilot implementation of a new Enterprise Resource Planning (ERP) system. Results from the study revealed that job stress was closely related to organizational commitment, job satisfaction, and psychological well-being, while change commitment was associated with higher organizational commitment and job satisfaction. Positive training reactions were linked to increased change commitment and organizational commitment, and change-specific self-efficacy also predicted commitment to change. Additionally, change self-efficacy and principal support significantly moderated the relationship between coping and organizational commi...

Research paper thumbnail of Trapping of Electrons around Nanoscale Metallic Wires Embedded in a Semiconductor Medium

arXiv: Mesoscale and Nanoscale Physics, 2019

We predict that conduction electrons in a semiconductor film containing a centered square array o... more We predict that conduction electrons in a semiconductor film containing a centered square array of metal nanowires normal to its plane are bound in quantum states around the central wires, if a positive bias voltage is applied between the wires at the square vertices and these latter. We obtain and discuss the eigenenergies and eigenfunctions of two models with different dimensions. The results show that the eigenstates can be grouped into different shells. The energy differences between the shells is typically a few tens of meV, which corresponds to frequencies of emitted or absorbed photons in a range of 3 THz to 20 THz approximately. These energy differences strongly depend on the bias voltage. We calculate the linear response of individual electrons on the ground level of our models to large-wavelength electromagnetic waves whose electric field is in the plane of the semiconductor film. The computed oscillator strengths are dominated by the transitions to the states in each shel...

Research paper thumbnail of A simple printed antenna with broadband property and omnidirectional radiation patterns of wire dipole

IEICE Electron. Express, 2020

This letter presents a unique printed dipole antenna which can roughly cover an 80% fractional ba... more This letter presents a unique printed dipole antenna which can roughly cover an 80% fractional bandwidth (VSWR  3), and especially possess the omnidirectional radiation patterns similar to those of wire dipole antenna at its all resonant modes. The antenna is simple with a microstrip-fed structure which it is just composed of two arms with each one forming the shape of a conventional "inset-fed" rectangular patch antenna. At the lowest operation frequency for VSWR = 3, the antenna size is only 0.3λ × 0.008λ, showing its effectiveness in many applications of small antennas.

Research paper thumbnail of Review, Analyze, and Design a Comprehensive Deep Reinforcement Learning Framework

ArXiv, 2020

Reinforcement learning (RL) has emerged as a standard approach for building an intelligent system... more Reinforcement learning (RL) has emerged as a standard approach for building an intelligent system, which involves multiple self-operated agents to collectively accomplish a designated task. More importantly, there has been a great attention to RL since the introduction of deep learning that essentially makes RL feasible to operate in high-dimensional environments. However, current research interests are diverted into different directions, such as multi-agent and multi-objective learning, and human-machine interactions. Therefore, in this paper, we propose a comprehensive software architecture that not only plays a vital role in designing a connect-the-dots deep RL architecture but also provides a guideline to develop a realistic RL application in a short time span. By inheriting the proposed architecture, software managers can foresee any challenges when designing a deep RL-based system. As a result, they can expedite the design process and actively control every stage of software d...

Research paper thumbnail of A Visual Communication Map for Multi-Agent Deep Reinforcement Learning

ArXiv, 2020

Multi-agent learning distinctly poses significant challenges in the effort to allocate a conceale... more Multi-agent learning distinctly poses significant challenges in the effort to allocate a concealed communication medium. Agents receive thorough knowledge from the medium to determine subsequent actions in a distributed nature. Apparently, the goal is to leverage the cooperation of multiple agents to achieve a designated objective efficiently. Recent studies typically combine a specialized neural network with reinforcement learning to enable communication between agents. This approach, however, limits the number of agents or necessitates the homogeneity of the system. In this paper, we have proposed a more scalable approach that not only deals with a great number of agents but also enables collaboration between dissimilar functional agents and compatibly combined with any deep reinforcement learning methods. Specifically, we create a global communication map to represent the status of each agent in the system visually. The visual map and the environmental state are fed to a shared-p...

Research paper thumbnail of A Prioritized objective actor-critic method for deep reinforcement learning

Neural Computing and Applications, 2021

An increasing number of complex problems have naturally posed significant challenges in decision-... more An increasing number of complex problems have naturally posed significant challenges in decision-making theory and reinforcement learning practices. These problems often involve multiple conflicting reward signals that inherently cause agents’ poor exploration in seeking a specific goal. In extreme cases, the agent gets stuck in a sub-optimal solution and starts behaving harmfully. To overcome such obstacles, we introduce two actor-critic deep reinforcement learning methods, namely Multi-Critic Single Policy (MCSP) and Single Critic Multi-Policy (SCMP), which can adjust agent behaviors to efficiently achieve a designated goal by adopting a weighted-sum scalarization of different objective functions. In particular, MCSP creates a human-centric policy that corresponds to a predefined priority weight of different objectives. Whereas, SCMP is capable of generating a mixed policy based on a set of priority weights, i.e., the generated policy uses the knowledge of different policies (each policy corresponds to a priority weight) to dynamically prioritize objectives in real time. We examine our methods by using the Asynchronous Advantage Actor-Critic (A3C) algorithm to utilize the multithreading mechanism for dynamically balancing training intensity of different policies into a single network. Finally, simulation results show that MCSP and SCMP significantly outperform A3C with respect to the mean of total rewards in two complex problems: Food Collector and Seaquest.

Research paper thumbnail of Manipulating Soft Tissues by Deep Reinforcement Learning for Autonomous Robotic Surgery

2019 IEEE International Systems Conference (SysCon), 2019

In robotic surgery, pattern cutting through a deformable material is a challenging research field... more In robotic surgery, pattern cutting through a deformable material is a challenging research field. The cutting procedure requires a robot to concurrently manipulate a scissor and a gripper to cut through a predefined contour trajectory on the deformable sheet. The gripper ensures the cutting accuracy by nailing a point on the sheet and continuously tensioning the pinch point to different directions while the scissor is in action. The goal is to find a pinch point and a corresponding tensioning policy to minimize damage to the material and increase cutting accuracy measured by the symmetric difference between the predefined contour and the cut contour. Previous study considers finding one fixed pinch point during the course of cutting, which is inaccurate and unsafe when the contour trajectory is complex. In this paper, we examine the soft tissue cutting task by using multiple pinch points, which imitates human operations while cutting. This approach, however, does not require the use of a multi-gripper robot. We use a deep reinforcement learning algorithm to find an optimal tensioning policy of a pinch point. Simulation results show that the multi-point approach outperforms the state-ofthe-art method in soft pattern cutting task with respect to both accuracy and reliability.

Research paper thumbnail of Application of St Gallen Categories in Predicting Survival for Patients With Breast Cancer in Vietnam

Cancer Control, 2019

Breast cancer is a heterogeneous disease with different tumor subtypes. Identifying risk categori... more Breast cancer is a heterogeneous disease with different tumor subtypes. Identifying risk categories will help make better treatment decisions. Hence, this study aimed to predict the survival outcomes of invasive breast cancer in Vietnam, using St Gallen 2007 classification. This study was conducted on 501 patients with breast cancer who had surgical operations, but had not received neoadjuvant chemotherapy, from 2011 to 2013. The clinicopathological characteristics were recorded. Immunohistochemistry staining was performed on ER, PR, HER2/neu, and Ki67 markers. For HER2/neu(2+), fluorescence in situ hybridization was used as the test. All patients with breast cancer were stratified according to 2007 St Gallen categories. Kaplan-Meier and log-rank models were used to analyze survival rates. There were 3.8% cases classified as low risk (LR), 72.1% as intermediate risk (IR1: 60.1% and IR2: 12.0%), and 24.1% as high risk (HR1: 11.8% and HR2: 12.3%). Patients who were LR had the best pro...

Research paper thumbnail of System Design Perspective for Human-Level Agents Using Deep Reinforcement Learning: A Survey

IEEE Access, 2017

Reinforcement learning (RL) has distinguished itself as a prominent learning method to augment th... more Reinforcement learning (RL) has distinguished itself as a prominent learning method to augment the efficacy of autonomous systems. Recent advances in deep learning studies have complemented existing RL methods and led to a crucial breakthrough in the effort of applying RL to automation and robotics. Artificial agents based on deep RL can take selective and intelligent actions comparable with those of a human to maximize the feedback reward from the interactive environment. In this paper, we survey recent developments in the literature regarding deep RL methods for building human-level agents. As a result, prominent studies that involve modeling every aspect of a human-level agent will be examined. We also provide an overview of constructing a framework for prospective autonomous systems. Moreover, various toolkits and frameworks are suggested to facilitate the development of deep RL methods. Finally, we open a discussion that potentially raises a range of future research directions in deep RL. INDEX TERMS Deep learning, human-level agents, reinforcement learning, robotics, survey, system design.

Research paper thumbnail of Investor confidence and mutual fund performance in emerging markets

Journal of Economic Studies, 2018

Purpose The purpose of this paper is to investigate the impact of investor confidence on mutual f... more Purpose The purpose of this paper is to investigate the impact of investor confidence on mutual fund performance in two relatively vulnerable but leading emerging markets, India and Pakistan. Design/methodology/approach A pooled ordinary least squared (OLS) model is used to look at two alternative measures of investor confidence and test for the relationship between investor confidence and mutual fund returns. To check the robustness of the findings, the authors also implement two-stage least squares and generalized method of moments techniques to control for unobserved heterogeneity, simultaneity and dynamic endogeneity problems in the regressors. Findings The paper finds that the returns of mutual funds are positively associated with investor confidence and an interaction effect exists between investor confidence and persistence in performance. The paper also confirms that returns from mutual funds are associated with different fund characteristics such as fund size, turnover, exp...