A5037771946- Face recognition and analysis
- AI in cancer detection
- Diabetic Foot Ulcer Assessment and Management
- Face and Expression Recognition
- Emotion and Mood Recognition
- Wound Healing and Treatments
- Cutaneous Melanoma Detection and Management
- Generative Adversarial Networks and Image Synthesis
- Radiomics and Machine Learning in Medical Imaging
- Advanced Image Processing Techniques
- Pressure Ulcer Prevention and Management
- Human Pose and Action Recognition
- Oral microbiology and periodontitis research
- Medical Image Segmentation Techniques
- Facial Nerve Paralysis Treatment and Research
- Image Retrieval and Classification Techniques
- Digital Imaging for Blood Diseases
- Nonmelanoma Skin Cancer Studies
- Gaze Tracking and Assistive Technology
- Advanced Neural Network Applications
- Facial Rejuvenation and Surgery Techniques
- Image and Signal Denoising Methods
- Gait Recognition and Analysis
- COVID-19 diagnosis using AI
- Balance, Gait, and Falls Prevention
Manchester Metropolitan University
2016-2025
Lancashire Teaching Hospitals NHS Foundation Trust
2018-2025
Hospital Punta Pacifica
2024
Malacca General Hospital
2024
University Malaya Medical Centre
2020
National Taiwan University of Science and Technology
2020
NIHR Oxford Musculoskeletal Biomedical Research Centre
2018
University of Bradford
1985-2013
Loughborough University
2006-2012
Bradford Royal Infirmary
1987
Breast lesion detection using ultrasound imaging is considered an important step of computer-aided diagnosis systems. Over the past decade, researchers have demonstrated possibilities to automate initial detection. However, lack a common dataset impedes research when comparing performance such algorithms. This paper proposes use deep learning approaches for breast and investigates three different methods: Patch-based LeNet, U-Net, transfer approach with pretrained FCN-AlexNet. Their compared...
Micro-facial expressions are spontaneous, involuntary movements of the face when a person experiences an emotion but attempts to hide their facial expression, most likely in high-stakes environment. Recently, research this field has grown popularity, however publicly available datasets micro-expressions have limitations due difficulty naturally inducing spontaneous micro-expressions. Other issues include lighting, low resolution and participant diversity. We present newly developed...
Early detection of skin cancer, particularly melanoma, is crucial to enable advanced treatment. Due the rapid growth in number cancers, there a growing need computerised analysis for lesions. The state-of-the-art public available datasets lesions are often accompanied with very limited amount segmentation ground truth labeling. Also, consist noisy expert annotations reflecting fact that precise represent boundary laborious and expensive. lesion vital locate accurately dermoscopic images...
Globally, in 2016, 1 out of 11 adults suffered from diabetes mellitus. Diabetic foot ulcers (DFU) are a major complication this disease, which if not managed properly can lead to amputation. Current clinical approaches DFU treatment rely on patient and clinician vigilance, has significant limitations, such as the high cost involved diagnosis, treatment, lengthy care DFU. We collected an extensive dataset images, contain different patients. In classification problem, we assessed two classes...
Current practice for diabetic foot ulcers (DFU) screening involves detection and localization by podiatrists. Existing automated solutions either focus on segmentation or classification. In this work, we design deep learning methods real-time DFU localization. To produce a robust model, collected an extensive database of 1775 images DFU. Two medical experts produced the ground truths data set outlining region interest with annotator software. Using five-fold cross-validation, overall, faster...
Recognition and analysis of Diabetic Foot Ulcers (DFU) using computerized methods is an emerging research area with the evolution image-based machine learning algorithms. Existing visual mainly focuses on recognition, detection, segmentation appearance DFU as well tissue classification. According to medical classification systems, presence infection (bacteria in wound) ischaemia (inadequate blood supply) has important clinical implications for assessment, which are used predict risk...
Various deep learning methods have been proposed to segment breast lesions from ultrasound images. However, similar intensity distributions, variable tumor morphologies and blurred boundaries present challenges for segmentation, especially malignant tumors with irregular shapes. Considering the complexity of images, we develop an adaptive attention U-net (AAU-net) automatically stably Specifically, introduce a hybrid module (HAAM), which mainly consists channel self-attention block spatial...
With recent advances in the field of deep learning, use convolutional neural networks (CNNs) medical imaging has become very encouraging. The aim our paper is to propose a patch-based CNN method for automated mass detection full-field digital mammograms (FFDM). In addition evaluating CNNs pretrained with ImageNet dataset, we investigate transfer learning particular domain adaptation. First, trained using large public database digitized (CBIS-DDSM dataset), and then model transferred tested...
The International Skin Imaging Collaboration (ISIC) datasets have become a leading repository for researchers in machine learning medical image analysis, especially the field of skin cancer detection and malignancy assessment. They contain tens thousands dermoscopic photographs together with gold-standard lesion diagnosis metadata. associated yearly challenges resulted major contributions to field, papers reporting measures well excess human experts. cancers can be divided into two groups -...
There has been a substantial amount of research involving computer methods and technology for the detection recognition diabetic foot ulcers (DFUs), but there is lack systematic comparisons state-of-the-art deep learning object frameworks applied to this problem. DFUC2020 provided participants with comprehensive dataset consisting 2,000 images training testing. This paper summarizes results by comparing learning-based algorithms proposed winning teams: Faster R–CNN, three variants R–CNN an...
Monitoring the healing progress of diabetic foot ulcers is a challenging process. Accurate segmentation can help podiatrists to quantitatively measure size wound regions assist prediction status. The main challenge in this field lack publicly available manual delineation, which be time consuming and laborious. Recently, methods based on deep learning have shown excellent results automatic medical images, however, they require large-scale datasets for training, there limited consensus perform...
Diabetic Foot Ulcer (DFU) is a major complication of Diabetes, which if not managed properly can lead to amputation. DFU appear anywhere on the foot and vary in size, colour, contrast depending various pathologies. Current clinical approaches treatment rely patients clinician vigilance, has significant limitations such as high cost involved diagnosis, lengthy care DFU. We introduce dataset 705 images. provide ground truth ulcer region surrounding skin that an important indicator for...
Automatic facial micro-expression (ME) analysis is a growing field of research that has gained much attention in the last five years. With many recent works testing on limited data, there need to spur better approaches are both robust and effective. This paper summarises 2nd Facial Micro-Expression Grand Challenge (MEGC 2019) held conjunction with 14th IEEE Conference Face Gesture Recognition (FG) 2019. In this workshop, we proposed challenges for two tasks- spotting recognition, aim...
Micro-expressions are brief spontaneous facial expressions that appear on a face when person conceals an emotion, making them different to normal in subtlety and duration. Currently, emotion classes within the CASME II dataset (Chinese Academy of Sciences Micro-expression II) based Action Units self-reports, creating conflicts during machine learning training. We will show classifying using Units, instead predicted removes potential bias human reporting. The proposed tested LBP-TOP (Local...
In recent years, the use of Convolutional Neural Networks (CNNs) in medical imaging has shown improved performance terms mass detection and classification compared to current state-of-the-art methods. This paper proposes a fully automated framework detect masses Full-Field Digital Mammograms (FFDM). is based on Faster Region-based Network (Faster-RCNN) model applied for detecting large-scale OPTIMAM Mammography Image Database (OMI-DB), which consists ∼80,000 FFDMs mainly from Hologic General...
Multistage processing of automated breast ultrasound lesions recognition is dependent on the performance prior stages. To improve current state art, we propose use end-to-end deep learning approaches using fully convolutional networks (FCNs), namely FCN-AlexNet, FCN-32s, FCN-16s, and FCN-8s for semantic segmentation lesions. We pretrained models based ImageNet transfer to overcome issue data deficiency. evaluate our results two datasets, which consist a total 113 malignant 356 benign assess...
The contributions of fiber atrophy, loss, in situ specific force, and voluntary activation to weakness sarcopenia remain unclear.To investigate, 40 older (20 women; age 72 ± 4 years) 31 younger adults (15 women, 22 3 completed measurements.The knee extensor maximal torque (MVC) was measured as well activation, patella tendon moment arm length, muscle volume, fascicle architecture estimate force.Fiber cross-sectional area (FCSA), numbers, connective tissue contents were also estimated from...
With the growth of popularity facial micro-expressions in recent years, demand for long videos with micro- and macro-expressions remains high. Extended from SAMM, a dataset released 2016, this paper presents SAMM Long Videos spontaneous recognition spotting. consists 147 343 159 micro-expressions. The is FACS-coded detailed Action Units (AUs). We compare our Chinese Academy Sciences Macro-Expressions Micro-Expressions (CAS(ME) <sup xmlns:mml="http://www.w3.org/1998/Math/MathML"...
This paper introduces the Diabetic Foot Ulcers dataset (DFUC2021) for analysis of pathology, focusing on infection and ischaemia. We describe data preparation DFUC2021 ground truth annotation, curation analysis. The final release consists 15,683 DFU patches, with 5,955 training, 5,734 testing 3,994 unlabeled patches. labels are four classes, i.e. control, infection, ischaemia both conditions. curate using image hashing techniques analyse separability UMAP projection. benchmark performance...
Chronic wounds and associated complications present ever growing burdens for clinics hospitals world wide. Venous, arterial, diabetic, pressure are becoming increasingly common globally. These conditions can result in highly debilitating repercussions those affected, with limb amputations increased mortality risk resulting from infection more common. New methods to assist clinicians chronic wound care therefore vital maintain high quality standards. This paper presents an improved HarDNet...
This paper proposes a novel approach to initial lesion detection in ultrasound breast images. The objective is automate the manual process of region interest (ROI) labeling computer‐aided diagnosis (CAD). We propose use hybrid filtering, multifractal processing, and thresholding segmentation automated ROI labeling. used 360 images evaluate performance proposed approach. Images were preprocessed using histogram equalization before filtering analysis conducted. Subsequently, was applied on...
We describe the development of a new mobile app called "FootSnap," to standardize photographs diabetic feet and test its reliability on different occasions between operators.FootSnap was developed by multidisciplinary team for use with iPad. The plantar surface 30 nondiabetic control were imaged using FootSnap two separate operators. Reproducibility foot images determined Jaccard similarity index (JSI).High intra- interoperator demonstrated JSI values 0.89-0.91 0.93-0.94 feet.Similarly high...