EYE-SCREEN-4-DPN: Development of an innovative Intelligent EYE imaging solution for SCREENing of Diabetic Peripheral Neuropathy
Lead Research Organisation:
Manchester Metropolitan University
Department Name: Ctr for Advanced Computational Science
Abstract
Abstracts are not currently available in GtR for all funded research. This is normally because the abstract was not required at the time of proposal submission, but may be because it included sensitive information such as personal details.
People |
ORCID iD |
| Liangxiu Han (Principal Investigator) |
| Title | Review and preliminary study |
| Description | The project only just started and we are currently conducting the literature review and preliminary study |
| Type Of Material | Technology assay or reagent |
| Year Produced | 2023 |
| Provided To Others? | No |
| Impact | n/a. The project only just started and we are currently conducting the literature review and preliminary study |
| Title | WAVEDIFFUR: A DIFFUSION SDE-BASED SOLVER FOR ULTRA MAGNIFICATION SUPER-RESOLUTION IN REMOTE SENSING IMAGES |
| Description | Deep neural networks have recently achieved significant advancements in remote sensing superresolu- tion (SR). However, most existing methods are limited to low magnification rates (e.g., ×2 or ×4) due to the escalating ill-posedness at higher magnification scales. To tackle this challenge, we redefine high-magnification SR as the ultra-resolution (UR) problem, reframing it as solving a conditional diffusion stochastic differential equation (SDE). In this context, we propose WaveDiffUR, a novel wavelet-domain diffusion UR solver that decomposes the UR process into sequential sub-processes addressing conditional wavelet components. WaveDiffUR iteratively reconstructs low-frequency wavelet details (ensuring global consistency) and high-frequency components (enhancing local fi- delity) by incorporating pre-trained SR models as plug-and-play modules. This modularity mitigates the ill-posedness of the SDE and ensures scalability across diverse applications. To address limitations in fixed boundary conditions at extreme magnifications, we introduce the cross-scale pyramid (CSP) constraint, a dynamic and adaptive framework that guides WaveDiffUR in generating fine-grained wavelet details, ensuring consistent and high-fidelity outputs even at extreme magnification rates. Extensive experiments demonstrate that the WaveDiffUR model, combined with CSP, achieves state-of-the-art performance by effectively minimizing degradation across key metrics: quantitative accuracy, perceptual quality, spectral consistency, and sharpness. Even when the magnification scale increases significantly from ×4 to ×128, the model maintains robust performance, with an average degradation of only 19.1%. At extreme magnifications (e.g., ×128), it outperforms benchmark mod- els, achieving up to 3 times the improvement in PSNR and SRE, showcasing superior image quality and spectral fidelity. By enabling robust ultra-resolution in remote sensing, WaveDiffUR opens new possibilities for applications in environmental monitoring, urban planning, disaster response, and precision agriculture. This study provides a significant step toward scalable, cost-effective, and high-fidelity solutions for real-world remote sensing challenges. |
| Type Of Material | Technology assay or reagent |
| Year Produced | 2025 |
| Provided To Others? | Yes |
| Impact | n/a |
| URL | https://arxiv.org/pdf/2412.18996 |
| Description | A key speaker --- Scalable Learning from Big data |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Professional Practitioners |
| Results and Impact | I was a key speaker at "The 3rd International Conference on Computing and Communication Networks (ICCCNet-2023)", Manchester Metropolitan University, Manchester, UK November 17-18th, 2023. |
| Year(s) Of Engagement Activity | 2023 |
| Description | BBC Northwest on our project report |
| Form Of Engagement Activity | A broadcast e.g. TV/radio/film/podcast (other than news/press) |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Professional Practitioners |
| Results and Impact | Dissemination our project via BBC northwest https://www.youtube.com/watch?v=kNosRndGK9g |
| Year(s) Of Engagement Activity | 2023 |
| URL | https://www.youtube.com/watch?v=kNosRndGK9g |
| Description | Keynotes speaker |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Professional Practitioners |
| Results and Impact | I have been invited to give a talk during IEEE international conferences.The 22nd IEEE International Conference on Trust, Security and Privacy in Computing and Communications (TrustCom-2023); The 17th IEEE International Conference on Big Data Science and Engineering (BigDataSE-2023); The 26th IEEE International Conference on Computational Science and Engineering (CSE-2023); The 21st IEEE International Conference on Embedded and Ubiquitous Computing (EUC-2023); The 11th IEEE International Conference on Smart City and Informatization . There are More than 500 audience, which sparked questions and discussions. |
| Year(s) Of Engagement Activity | 2023 |
| Description | Scalable and Explainable Learning for Alzheimer's Disease Diagnosis with Structural MRI |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | Regional |
| Primary Audience | Professional Practitioners |
| Results and Impact | Liangxiu Han gave an invited speaker for a workshop by HarmonicAI( a research project funded by Horizon Europe Marie Sklodowska-Curie Actions Staff Exchange scheme and UKRI). The workshop will integrate human-centred co-design approaches with systems thinking, aiming to: • Identify barriers to the acceptance and adoption of AI-powered digital health services. • Understand healthcare stakeholders' expectations in the key areas of explainability, fairness, and privacy. The event was attended by approximately 35 to 50 participants, and the talk generated engaging discussions and questions, contributing to the project's next steps. |
| Year(s) Of Engagement Activity | 2025 |