A5037238008- Medical Image Segmentation Techniques
- Advanced Steganography and Watermarking Techniques
- Chaos-based Image/Signal Encryption
- AI in cancer detection
- Digital Media Forensic Detection
- Remote-Sensing Image Classification
- Brain Tumor Detection and Classification
- Advanced Image and Video Retrieval Techniques
- Remote Sensing and Land Use
- Medical Imaging and Analysis
- Advanced Data Compression Techniques
- Radiomics and Machine Learning in Medical Imaging
- Face and Expression Recognition
- Face recognition and analysis
- Advanced Neural Network Applications
- Image Retrieval and Classification Techniques
- Reproductive System and Pregnancy
- Experimental Learning in Engineering
- Simulation and Modeling Applications
- Robotics and Sensor-Based Localization
- Visual Attention and Saliency Detection
- Teaching and Learning Programming
- Advanced Image Fusion Techniques
- Infectious Encephalopathies and Encephalitis
- Innovative Educational Techniques
Tianjin Chengjian University
2010-2025
Jinan City People's Hospital
2023
Shandong First Medical University
2023
Qilu Hospital of Shandong University
2023
Shandong Center for Disease Control and Prevention
2023
Fort Valley State University
2017-2021
University of the Ryukyus University Hospital
2020
Valley City State University
2020
Binzhou People's Hospital
2015-2017
Ritsumeikan University
2013-2016
Acute necrotizing encephalopathy (ANE) is a devastating neurologic condition that can arise following variety of systemic infections, including influenza and SARS-Cov-2. The clinical features COVID-19-associated ANE in pediatric patients based on multi-case data have not yet been described remain obscure. We reviewed 12 to better describe the with COVID-19. retrospectively collected summarized children Clinical were from children, their general status, symptoms, laboratory tests,...
It is widely known that morphological changes of the liver and spleen occur during clinical course chronic diseases. In this paper, we proposed a analysis method based on statistical shape models (SSMs) for computer-aided diagnosis quantification liver. We constructed not only SSM but also joint morphologic cirrhotic in CT images. The effective modes are selected both its accumulation contribution rate correlation with doctor’s opinions (stage labels). then learn mapping function between...
Random walks-based (RW) segmentation methods have been proven to a potential application in segmenting the medical image with minimal interactive guidance. However, approach leads large-scale graphs due number of nodes equal voxel number. Also, is inaccurate because unavailability appropriate initial seed points. It challenge use RW-based algorithm segment organ regions from 3D images interactively. In this paper, knowledge-based framework for multiple organs proposed based on random walks....
Objective: To investigate the clinical significance of human epididymal secretory protein E4 (HE4) in combination with cancer antigen 125 (CA125) diagnosis endometrial cancer.Methods: One hundred and fifty patients who were admitted to Binzhou People’s Hospital, Shandong, China, between June 2013 July 2014, enrolled set as an group; another one benign uterine diseases healthy females also enrolled. The serum was collected from subjects for detection HE4 level. level CA125 detected using...
Medical images and data may encounter tampering, compromising patient's privacy other safety issues in the process of storage transmission information. This paper proposes an algorithm zero watermarking to address these using DFT. The avoids sophisticated finding Region Non-Interest (RONI) medical images, uses a part sign sequence DFT coefficients as feature vector for enhancing robustness. Even if is public, it can still withstand malicious attacks by private key protect image. Experimental...
Random walk (RW) method has been widely used to segment the organ in volumetric medical image. However, it leads a very large-scale graph due number of nodes equal voxel and inaccurate segmentation because unavailability appropriate initial seed point setting. In addition, classical RW algorithm was designed for user mark few pixels with an arbitrary labels, regardless intensity shape information organ. Hence, we propose prior knowledge-based Bayes random framework image slice-by-slice...
When medical images transmitted and stored in hospitals, it require strict security, confidentiality integrity. However, the transmission of wireless wired networks has made information vulnerable to attacks like tampering, hacking etc. And ROI image is unable tolerate significant changes. In order dealing these problems, we have proposed an algorithm that introducing digital watermarking technology increase security images. The scheme uses a part sign sequence DCT coefficients as feature...
In the watermarking algorithm of medical image, principal problems faced are how to determine region interest (ROI) and improve hiding capacity. This paper proposes a robust multiple watermarks for image based on DWT DCT, which can effectively solve referred problems. The combines visual feature vector images, encryption technology with third party authentication, avoids tedious process selecting ROI. enhance security, confidentiality integrity in application clinical. results experiment...
Sparse representation classification (SRC) is being widely applied to target detection in hyperspectral images (HSI). However, due the problem HSI that high-dimensional data contain redundant information, SRC methods may fail achieve high performance, even with a large number of spectral bands. Selecting subset predictive features space an important and challenging for image classification. In this paper, we propose novel discriminant feature learning (DFL) method, which combines spatial...
Ubiquitous computing is an important technology in medicine that predicted to support doctors anywhere and anytime. To help achieve it, this paper develops the Interactive Segmentation Visualization System for Medical images on Mobile devices (ISVS_M2), which originally was designed work workstations, but also works a wide range of mobile via client-server platform. The developed ISVS_M2 basically consists three modules: asegmentation module implemented server; commutation modules both...
A random walks-based (RW) segmentation method has been gaining popularity in recent years with its ability to interactively segment the objects minimal guidance. It potential applications segmenting 3D image. However, due large computational burden of classical RW algorithm, it is a challenge use this algorithm medical images interactively. Hence, knowledge-based framework for liver proposed based on walks and narrow band threshold (RWNBT). Our strategy employ previous segmented slice...
Medical image data require strict security, confidentiality and integrity when transmitted between hospitals. The involvement of wireless transmission has made the medical vulnerable to attacks like tampering, hacking etc. To dealing challenges we have proposed a algorithm that introducing digital watermarking technology increase security images they through network. scheme uses part sign sequence DWT-DCT coefficients as feature vector for enhancing robustness against common geometric...
Detection of the face region is a key step in recognition system. Thermal images are widely used many applications where normal visibility reduced, impaired or ineffective, such as night surveillance and fugitive searches. However, low spatial resolution brings challenges to detection thermal images. Viola-Jones an object method for detection. The algorithm suffers from missed faces wrongly detected non-face objects due To improve performance images, we propose incorporate cross-examples...
For the copyright protection of digital video, a new video watermarking algorithm based on 3D-DWT&DCT is proposed. We select segment and get coefficients using 3D-DWT&DCT. The sign sequence 3DDWT&DCT used as feature vector, which utilized to enhance robustness against geometrical attacks. Combining concept third-party, we describe how embed extract multiple watermarks. Simulation results demonstrate proposed has strong
Medical images must be stored and translated in a secure way. This paper proposes multiple watermarks technique based on medical image features to serve these purposes. A part of sign sequence DWT DFT coefficients are used as feature vector for enhancing the robustness against rotation, scaling, translation cropping attacks. In contrast conventional approaches, our method can strengthen fidelity through considering local characteristics, without sacrificing quality images. We describe how...
Locoregional therapy (LT) is a popular method to cure hepatocellular carcinoma (HCC) of liver in the last few years. Evaluating treated margin critical factor for predicting local recurrence after locoregional treatment done. In this paper, we proposed non-rigid registration with constraint anatomic landmarks assessing therapy. Compared conventional registration, which uses only mutual information as cost function (without any constraints anatomical structure), our can achieve high accuracy...