A5031252766- Retinal Imaging and Analysis
- Computational Drug Discovery Methods
- Glaucoma and retinal disorders
- Surgical Simulation and Training
- Medical Imaging and Analysis
- Retinal Diseases and Treatments
- Bioinformatics and Genomic Networks
- Corneal surgery and disorders
- Advanced Vision and Imaging
- Intraocular Surgery and Lenses
- Biosimilars and Bioanalytical Methods
- Topological and Geometric Data Analysis
- Retinal and Optic Conditions
- Artificial Intelligence in Healthcare and Education
- Robotics and Sensor-Based Localization
- Ocular Surface and Contact Lens
- Protein Structure and Dynamics
- Image Retrieval and Classification Techniques
- 3D Surveying and Cultural Heritage
- Domain Adaptation and Few-Shot Learning
- Digital Imaging in Medicine
- Advanced Neural Network Applications
- Tuberculosis Research and Epidemiology
Southern University of Science and Technology
2022-2024
University of Birmingham
2022-2023
Quality degradation (QD) is common in the fundus images collected from clinical environment. Although diagnosis models based on convolutional neural networks (CNN) have been extensively used to interpret retinal images, their performances under QD not assessed. To understand effects of performance CNN-based model, a systematical study proposed this paper. In our study, controlled by independently or simultaneously importing quantified interferences (e.g., image blurring, artifacts, and light...
Drug combinations could trigger pharmacological therapeutic effects (TEs) and adverse (AEs). Many computational methods have been developed to predict TEs, e.g. the synergy scores of anti-cancer drug combinations, or AEs from drug-drug interactions. However, most treated TEs predictions as two separate tasks, ignoring potential mechanistic commonalities shared between them. Based on previous clinical observations, we hypothesized that by learning learn underlying MoAs (mechanisms actions)...
Abstract Motivation To predict drug targets, graph-based machine-learning methods have been widely used to capture the relationships between drug, target and disease entities in drug–disease–target (DDT) networks. However, many cannot explicitly consider types at inference time so will same for a given under any condition. Meanwhile, DDT networks are usually organized hierarchically carrying interactive involved entities, but these methods, especially those based on Euclidean embedding fully...
To study the association between dynamic iris change and primary angle-closure disease (PACD) with anterior segment optical coherence tomography (AS-OCT) videos develop an automated deep learning system for screening as well validate its performance.A total of 369 AS-OCT (19,940 frames)-159 subjects 210 normal controls (two datasets using different capturing devices)-were included. The correlation changes (pupil constriction) PACD was analyzed based on clinical parameters diameter) under...
In recent years, visual-inertial Simultaneous Localization and Mapping (SLAM) techniques have achieved excellent results due to the complementary nature of visual inertial sensors, challenges remain in some environments. When encountering a challenging environment, image features become an important factor affecting success rate accuracy localization SLAM system. Due Convolutional Neural Network (CNN) has advantage extracting effective more accurately, this paper makes new attempt replace...
Semantic segmentation of surgery scenarios is a fundamental task for computer-aided systems. Precise surgical instruments and anatomies contributes to capturing accurate spatial information tracking. However, uneven reflection class imbalance lead the in cataract challenging task. To desirably conduct segmentation, network with multi-view decoders (MVD-Net) proposed present generalizable surgery. Two discrepant are implemented achieve learning backbone U-Net. The experiment carried out on...
Motivation: To predict novel drug targets, graph-based machine learning methods have been widely used to capture the relationships between drug, target, and disease entities in heterogeneous biological networks. However, most of existing cannot explicitly consider types when predicting targets. More importantly, drug-disease-target (DDT) networks could exhibit multi-relational hierarchical sub-structures with information interactive functions, but these methods, especially those based on...
Accurate surgical video semantic segmentation is vital for computer-aided surgery. Semi-supervised algorithms produce pseudo labels to solve the problem of lack labels, as it very difficult obtain pixel-level from doctors or researchers. However, most consider videos independent images, which cannot some issues caused by complex surgery scenarios, such blurred instruments. The paper proposes a novel Cross Supervision Inter-frame (CSI) method using inter-frame information crosswise supervise...
Recognition and localization of surgical detailed actions is an essential component developing a context-aware decision support system. However, most existing detection algorithms fail to provide high-accuracy action classes even having their locations, as they do not consider the surgery procedure's regularity in whole video. This limitation hinders application. Moreover, implementing predictions clinical applications seriously needs convey model confidence earn entrustment, which...
Background. Precise and comprehensive characterizations from anterior segment optical coherence tomography (AS-OCT) are of great importance in facilitating the diagnosis angle-closure glaucoma. Existing automated analysis methods focus on analyzing structural properties identified single AS-OCT image, which is limited to comprehensively representing status chamber angle (ACA). Dynamic iris changes evidenced as a risk factor primary Method. In this work, we detecting ACA videos, captured...