A5057905491- Retinal Imaging and Analysis
- Glaucoma and retinal disorders
- Retinal Diseases and Treatments
- Optical Coherence Tomography Applications
- Retinal and Optic Conditions
- Medical Image Segmentation Techniques
- Digital Imaging for Blood Diseases
- 3D Shape Modeling and Analysis
- Ocular Surface and Contact Lens
- Corneal surgery and disorders
- Visual Attention and Saliency Detection
- Computer Graphics and Visualization Techniques
- COVID-19 diagnosis using AI
- Advanced Vision and Imaging
- Anomaly Detection Techniques and Applications
- AI in cancer detection
- Robotics and Sensor-Based Localization
- Medical Imaging and Analysis
- Cerebrovascular and Carotid Artery Diseases
- Advanced Neural Network Applications
- Radiomics and Machine Learning in Medical Imaging
- Advanced Image and Video Retrieval Techniques
- Video Surveillance and Tracking Methods
- Image Retrieval and Classification Techniques
- Advanced Image Processing Techniques
Ningbo Institute of Industrial Technology
2017-2025
Chinese Academy of Sciences
2017-2025
Wenzhou Medical University
2022-2025
University of Liverpool
2015-2024
Nanjing University of Posts and Telecommunications
2023-2024
Zhejiang Pharmaceutical College
2024
Shanghai University of Traditional Chinese Medicine
2019-2023
Affiliated Eye Hospital of Wenzhou Medical College
2022-2023
State Key Laboratory of Drug Research
2023
Shanghai Institute of Materia Medica
2020-2023
Medical image segmentation is an important step in medical analysis. With the rapid development of a convolutional neural network processing, deep learning has been used for segmentation, such as optic disc blood vessel detection, lung cell and so on. Previously, U-net based approaches have proposed. However, consecutive pooling strided operations led to loss some spatial information. In this paper, we propose context encoder (CE-Net) capture more high-level information preserve 2D...
Automated detection of blood vessel structures is becoming crucial interest for better management vascular disease. In this paper, we propose a new infinite active contour model that uses hybrid region information the image to approach problem. More specifically, an perimeter regularizer, provided by using L(2) Lebesgue measure γ -neighborhood boundaries, allows small oscillatory (branching) than traditional models based on length feature's boundaries (i.e., H(1) Hausdorff measure)....
Optical Coherence Tomography Angiography (OCTA) is a non-invasive imaging technique that has been increasingly used to image the retinal vasculature at capillary level resolution. However, automated segmentation of vessels in OCTA under-studied due various challenges such as low visibility and high vessel complexity, despite its significance understanding many vision-related diseases. In addition, there no publicly available dataset with manually graded for training validation algorithms. To...
Multiple instance learning (MIL) has been increasingly used in the classification of histopathology whole slide images (WSIs). However, MIL approaches for this specific problem still face unique challenges, particularly those related to small sample cohorts. In these, there are limited number WSI slides (bags), while resolution a single is huge, which leads large patches (instances) cropped from slide. To address issue, we propose virtually enlarge bags by introducing concept pseudo-bags, on...
The development of medical imaging techniques has greatly supported clinical decision making. However, poor quality, such as non-uniform illumination or imbalanced intensity, brings challenges for automated screening, analysis and diagnosis diseases. Previously, bi-directional GANs (e.g., CycleGAN), have been proposed to improve the quality input images without requirement paired images. these methods focus on global appearance, imposing constraints structure illumination, which are...
The detection of retinal vessel is great importance in the diagnosis and treatment many ocular diseases. Many methods have been proposed for detection. However, most algorithms neglect connectivity vessels, which plays an important role diagnosis. In this paper, we propose a novel method includes dense dilated network to get initial vessels probability regularized walk algorithm address fracture issue integrates newly feature extraction blocks into encoder-decoder structure extract...
Abstract Aims/hypothesis Corneal confocal microscopy is a rapid non-invasive ophthalmic imaging technique that identifies peripheral and central neurodegenerative disease. Quantification of corneal sub-basal nerve plexus morphology, however, requires either time-consuming manual annotation or less-sensitive automated image analysis approach. We aimed to develop validate an artificial intelligence-based, deep learning algorithm for the quantification fibre properties relevant diagnosis...
Alzheimer's disease (AD) accounts for 60%-70% of all dementia cases, and clinical diagnosis at its early stage is extremely difficult. As several new drugs aiming to modify progression or alleviate symptoms are being developed, assess their efficacy, novel robust biomarkers brain function urgently required. This paper aims explore a routine gain such using the quantitative analysis electroencephalography (QEEG). proposes supervised classification framework that uses EEG signals classify...
Automated detection of vascular structures is great importance in understanding the mechanism, diagnosis, and treatment many pathologies. However, automatic continues to be an open issue because difficulties posed by multiple factors, such as poor contrast, inhomogeneous backgrounds, anatomical variations, presence noise during image acquisition. In this paper, we propose a novel 2-D/3-D symmetry filter tackle these challenging issues for enhancing vessels from different imaging modalities....
Our application concerns the automated detection of vessels in retinal images to improve understanding disease mechanism, diagnosis and treatment a number systemic diseases. We propose new framework for segmenting vasculatures with much improved accuracy efficiency. The proposed consists three technical components: Retinex-based image inhomogeneity correction, local phase-based vessel enhancement graph cut-based active contour segmentation. These procedures are applied following order....
Semi-supervised approaches for crowd counting attract attention, as the fully supervised paradigm is expensive and laborious due to its request a large number of images dense scenarios their annotations. This paper proposes spatial uncertainty-aware semi-supervised approach via regularized surrogate task (binary segmentation) problems. Different from existing learning-based methods, exploit unlabeled data, our proposed teacher-student framework focuses on high confident regions' information...
With the development of convolutional neural network, deep learning has shown its success for retinal disease detection from optical coherence tomography (OCT) images. However, often relies on large scale labelled data training, which is oftentimes challenging especially with low occurrence. Moreover, a system trained data-set one or few diseases unable to detect other unseen diseases, limits practical usage in screening. To address limitation, we propose novel anomaly framework termed...
Segmentation is a fundamental task in biomedical image analysis. Unlike the existing region-based dense pixel classification methods or boundary-based polygon regression methods, we build novel graph neural network (GNN) based deep learning framework with multiple reasoning modules to explicitly leverage both region and boundary features an end-to-end manner. The mechanism extracts discriminative features, referred as initialized node embeddings, using proposed Attention Enhancement Module...
Though deep learning has shown successful performance in classifying the label and severity stage of certain diseases, most them give few explanations on how to make predictions. Inspired by Koch's Postulates, foundation evidence-based medicine (EBM) identify pathogen, we propose exploit interpretability application medical diagnosis. By isolating neuron activation patterns from a diabetic retinopathy (DR) detector visualizing them, can determine symptoms that DR identifies as evidence...
Cataracts are the leading cause of vision loss worldwide. Restoration algorithms developed to improve readability cataract fundus images in order increase certainty diagnosis and treatment for patients. Unfortunately, requirement annotation limits application these clinics. This paper proposes a network annotation-freely restore cataractous (ArcNet) so as boost clinical practicability restoration. Annotations unnecessary ArcNet, where high-frequency component is extracted from replace...
Background Optical coherence tomography angiography (OCTA) enables fast and non-invasive high-resolution imaging of retinal microvasculature is suggested as a potential tool in the early detection microvascular changes Alzheimer’s Disease (AD). We developed standardised OCTA analysis framework compared their extracted parameters among controls AD/mild cognitive impairment (MCI) cross-section study. Methods defined geometrical at different layers foveal avascular zone (FAZ) from segmented...
Cerebrovascular diseases (CVDs) remain a leading cause of global disability and mortality. Digital Subtraction Angiography (DSA) sequences, recognized as the gold standard for diagnosing CVDs, can clearly visualize dynamic flow reveal pathological conditions within cerebrovasculature. Therefore, precise segmentation cerebral arteries (CAs) classification between their main trunks branches are crucial physicians to accurately quantify diseases. However, achieving accurate CA in DSA sequences...
Leakage in retinal angiography currently is a key feature for confirming the activities of lesions management wide range diseases, such as diabetic maculopathy and paediatric malarial retinopathy. This paper proposes new saliency-based method detection leakage fluorescein angiography. A superpixel approach firstly employed to divide image into meaningful patches (or superpixels) at different levels. Two saliency cues, intensity compactness, are then proposed estimation map each individual...
Apart from the need for superior accuracy, healthcare applications of intelligent systems also demand deployment interpretable machine learning models which allow clinicians to interrogate and validate extracted medical knowledge. Fuzzy rule-based are generally considered that able reflect associations between conditions associated symptoms, through use linguistic if-then statements. Systems built on top fuzzy sets particular appealing since they enable tolerance vague imprecise concepts...
The choroid provides oxygen and nourishment to the outer retina thus is related pathology of various ocular diseases. Optical coherence tomography (OCT) advantageous in visualizing quantifying vivo. However, its application study still limited for two reasons. (1) lower boundary (choroid-sclera interface) OCT fuzzy, which makes automatic segmentation difficult inaccurate. (2) visualization hindered by vessel shadows from superficial layers inner retina. In this paper, we propose incorporate...