A5102890535- Radiomics and Machine Learning in Medical Imaging
- Head and Neck Cancer Studies
- Advanced X-ray and CT Imaging
- Colorectal Cancer Screening and Detection
- Medical Imaging Techniques and Applications
- Advanced Image and Video Retrieval Techniques
- Image Retrieval and Classification Techniques
- Topic Modeling
- Ear and Head Tumors
- Robotic Path Planning Algorithms
- Nonmelanoma Skin Cancer Studies
- Artificial Intelligence in Healthcare and Education
- Cutaneous Melanoma Detection and Management
- Natural Language Processing Techniques
- Robotics and Sensor-Based Localization
- Aerospace Engineering and Control Systems
- Free Radicals and Antioxidants
- Stroke Rehabilitation and Recovery
- Iron oxide chemistry and applications
- Infrared Target Detection Methodologies
- Hand Gesture Recognition Systems
- Fire Detection and Safety Systems
- Economic and Business Development Strategies
- Handwritten Text Recognition Techniques
- Thermography and Photoacoustic Techniques
FPT University
2023-2024
Information Technology Institute
2020-2024
Vietnam Academy of Science and Technology
2023
Le Quy Don Technical University
2023
Ministry of Health
2023
European Organisation for Research and Treatment of Cancer
1995-2021
Health Affairs
2021
Institute of Mathematics
2020
London Health Sciences Centre
2020
Ton Duc Thang University
2019
Colonoscopy is considered the gold-standard investigation for colorectal cancer screening. However, polyps miss rate in clinical practice relatively high due to different factors. This presents an opportunity use AI models automatically detect and segment polyps, supporting clinicians reduce number of missed. Inspired by success UNets, a popular strategy solving medical image segmentation tasks, this article proposes novel framework polyp called CRF-EfficientUNet, which enhances UNet using...
We propose a computationally efficient and effective novel neural network for predicting the next-day's closing price of US stocks in different sectors: technology, energy finance. In this paper we used functional link artificial (FLANN) making stock prediction. modeled trend movement as dynamic system apply FLANN to predict behavior. addition historical pricing data, considered other financial indicators such industrial indices technical indicators, better accuracy. showed its superior...
Automatic polyp detection and segmentation are highly desirable for colon screening due to miss rate by physicians during colonoscopy, which is about 25%. Diagnosis of polyps in colonoscopy videos a challenging task variations the size shape polyps. In this paper, we adapt U-net evaluate its performance with different modern convolutional neural networks as encoder segmentation. One major challenges training raises when data unbalanced, pixels often much lower numbers than non-polyp pixels....
Attenuation correction (AC) images are essential to perform qualitative and quantitative analysis in myocardial perfusion tomography (MPI) using a single-photon emission (SPECT) scanner. Hybrid SPECT/CT systems were born by combination of CT scanner regular SPECT machine directly measure photon degradation the body. However, reality, these often very expensive not available most health centers Vietnam. In this paper, we propose method filtering attenuated noise 3DUnet-Gan network....
This paper primarily focuses on the relationship between transformational leadership style, employee innovation, and job performance, with a particular emphasis its significance within context of startup enterprises. The authors point out that while existing research has extensively examined validated positive influence performance creativity (as evidenced by studies such as Naguib & Naem, 2018; Ibrahim et al., 2023; Shang, 2023), empirical evidence regarding this in specific environment...
Breast cancer is a prominent cause of death among women worldwide. Infrared thermography, due to its cost-effectiveness and non-ionizing radiation, has emerged as promising tool for early breast diagnosis. This article presents hybrid model approach detection using thermography images, designed process classify these images into healthy or cancerous categories, thus supporting disease Multiple pre-trained convolutional neural networks are employed image feature extraction, filter methods...
The process of searching for information to serve the construction operational plan documents on a digital map is still being done manually and needs be automated in order improve efficiency. Speech recognition natural language processing technologies, commonly used chatbots, virtual assistants, voice commands, search, could promising tools overcome this problem. This paper proposes framework deploying search engine that uses Whisper, deep learning-based automatic speech model, combines...
Abstract Arsenic-polluted water is a global concern and puts millions of people at risk developing cancer. The improvement aqueous solution coexisting with arsenite arsenate using iron mixed porous clay pellets was investigated in batch fixed-bed column systems. Batch studies showed that the removal rate occurred two main phases an equilibrium time 52 h. pseudo-second-order model well described experimental data. Isotherm data were fitted by Langmuir–Freundlich model. efficiency...
Abstract This paper presents a two‐stage deep learning framework, RR‐HCL‐SVM, designed to aid in the assessment of residual thyroid tissues following thyroidectomy, utilizing single‐photon emission computed tomography (SPECT) images. Leveraging power learning, our model offers comprehensive solution for detection and remaining tissues. To enhance accuracy, we introduce unique combination features, incorporating Radio Scan Index (RSI) radiomics features. These features not only improve...
This article presents an innovative approach to ascertain the most effective ablation dosages for thyroid cancer treatment following thyroidectomy. The methodology utilizes Decision Trees and places significant emphasis on interpretability of medical decision-making. By incorporating clinical data Radioactive Scan Index (RSI) into Tree algorithms, our offers transparent planning insights. means a case study, we illustrate function in clarifying pivotal elements that impact dosage...
This paper presents a novel model for extractive summarization that integrates context representation from pre-trained language (PLM), such as BERT, with prior knowledge derived unsupervised learning methods. Sentence importance assessment is crucial in summarization, providing indicators of sentence within document. Our introduces method estimating based on knowledge, complementing the contextual offered by PLMs like BERT. Unlike previous approaches primarily relied alone, our leverages...
In GNSS-denied environments, accurate Unmanned Aerial (UAV) localization faces significant challenges. This paper introduces a vision-based method combining autoencoder and SIFT algorithms, referred to as AE+SIFT. The compresses high-resolution map images into low-dimensional vectors, which are stored in database for efficient retrieval. During the process, UAV encoded matched with database, followed by homography projection precise positioning. AE+SIFT approach enhances accuracy, achieving...
Finding plagiarism strings between two given documents are the main task of detection problem. Traditional approaches based on string matching not very useful in cases similar semantic plagiarism. Deep learning solve this problem by measuring similarity pairs sentences. However, these still face following challenging points. First, it is impossible to where only part a sentence belongs passage. Second, sentential without considering context surrounding sentences leads decreasing accuracy. To...
Colonoscopy image classification is an task that predicts whether colonoscopy images contain polyps or not. It important input for automatic polyp detection system. Recently, deep neural networks have been widely used due to the feature extraction with high accuracy. However, training these requires a large amount of manually annotated data, which expensive acquire and limited by available resources endoscopy specialists. We propose novel method using self-supervised visual learning overcome...
Coronary artery disease (CAD) is one of the most common pathological conditions and major global cause death. Myocardial perfusion imaging (MPI) using single-photon emission computed tomography (SPECT) a non-invasive method plays an essential role in diagnosing CAD. However, there currently shortage doctors who can diagnose SPECT-MPI developing countries, especially Vietnam. Research on deploying machine learning deep supporting CAD diagnosis has been noticed for long time. these methods...