Topic 1: Using Advanced Technology to Address Health Care Challenges
Project Title: Embracing Artificial Intelligence-Assisted Upper and Lower Urinary Tract Assessment in Primary Care and Nurse Clinic Settings
Project Coordinator: Prof Jeremy Yuen-chun Teoh (CUHK)
Abstract
Haematuria, i.e. blood in urine, is a common urological presentation, and it has been one of the leading causes of referrals to the Urology Specialist Outpatient Clinic. Usually, patients will require haematuria workup including ultrasound scan for the upper urinary tract (i.e. kidneys and ureters) and flexible cystoscopy for the lower urinary tract (i.e. bladder, prostate and urethra), mainly to rule out urinary stone disease and urological malignancies. With an ageing population in Hong Kong, we are witnessing an increasing demand in managing patients with haematuria, resulting in manpower issues and long waiting time for every single step of the haematuria diagnostic pathway. In our hospital, the average waiting time is around 120 weeks for the first Urology Specialist Outpatient Clinic, 52 weeks for an ultrasound scan of the urinary system, and 16 weeks for a flexible cystoscopy. We are in need of an innovative, highly efficient and easily adoptable way of managing this common yet relatively simple urological presentation.
In this project, we aim to change the whole diagnostic pathway of haematuria by providing a point-of-care solution that is applicable in both primary care and nurse clinic settings. For the upper urinary tract, ultrasound is a simple imaging modality that can be performed bedside, and by developing an AI model with excellent diagnostic performance, a comprehensive assessment of the upper urinary tract can be achieved even in the primary care and nurse clinic settings. For the lower urinary tract, we propose to develop a novel urinary catheter with a built-in camera that allows visualisation of the lower urinary tract, supplemented by an AI model that can provide real-time cystoscopic assessment; a comprehensive assessment of the lower urinary tract can thus be achieved in a simple clinic setting. We will conduct a pilot study to implement the above novel technology in both primary care and nurse clinic settings. We strongly believe that such novel approach and technology will be able to empower primary care physicians and nurses to have an accurate and rapid point-of-care assessment of this common and important urological problem, and it will have significant impact in reducing the workload in the Urology Specialist Outpatient Clinics as well as the public hospitals.
Topic 1: Using Advanced Technology to Address Health Care Challenges
Project Title: An Integrated Technology Platform for Next-generation Cancer Immunotherapy - from Identification of Tumor Neoantigens to Development of Novel Therapeutic Vaccine Modalities
Project Coordinator: Prof Yanxiang Zhao (PolyU)
Abstract
Cancer immunotherapy harnesses the power of host immunity to eliminate tumor cells and has transformed the landscape of cancer treatment. Prominent examples such as antibodies that act as immune checkpoint inhibitors (ICIs) and chimeric antigen receptor-T cell therapy (CAR-T) have achieved remarkable clinical success, particularly by enabling long-term survival in some patients. Nonetheless, immunotherapies have significant limitations. Many cancer types are refractory to ICIs. CAR-T therapy is mostly effective for blood cancers but not solid tumors.
Neoantigen-based therapeutic vaccine has emerged as a promising new modality for cancer immunotherapy. This approach involves administering tumor neoantigens as vaccines to patients to elicit specific and durable immune response. Some leading candidates, particularly mRNA-based vaccines, have shown encouraging results in early-stage clinical trials. However, major challenges remain to be addressed for this approach, particularly regarding the insufficient immunogenicity of neoantigens and the immunosuppressive tumor microenvironment.
To tackle these challenges and seize the enormous opportunities presented by next-generation cancer immunotherapy, our research team proposes to build an integrated technology platform to develop Peptide-based Immunogenic Neoantigen Vaccines (PIN-Vax). This platform consists of four interconnected modules that collectively cover the full preclinical development cycle. Module 1 is an AI-backed multi-omics platform to discover shared and personalized tumor neoantigens from clinical samples. Module 2 is a LipoNeoAg system that uses synthetic biology tools to add lipid moieties to peptide-based neoantigens to enhance their immunogenicity. Module 3 is an Autotide system that uses autophagy-targeting stapled peptides as both anti-proliferative agents and novel immune modulators. Module 4 is a preclinical evaluation system to assess the anti-tumor efficacy of potential PIN-Vax candidates using the LipoNeoAg+Autotide format. We reason this novel format may elicit stronger anti-tumor immunity than the neoantigen-only format currently undergoing clinical trials.
To demonstrate the feasibility of our PIN-Vax platform, we plan to apply it to HPV-associated cervical cancer and HBV-associated hepatocellular carcinoma as both cancer types contain virus- derived neoantigens suitable for vaccine development. We aim to develop a rich pipeline of PIN- Vax candidates and assess their anti-tumor efficacy. We also plan to combine PIN-Vax candidates with ICIs for further synergistic effect.
We have assembled an interdisciplinary team of academic researchers, clinicians, and industry collaborators to build this PIN-Vax platform. Our track record and preliminary studies ensure the feasibility of this project. Our long-term goal is to develop this platform into an innovative engine of next-generation cancer immunotherapy to benefit cancer patients in Hong Kong and worldwide.
Topic 2: Striving towards Carbon Neutrality before 2050
Project Title: Coastal Blue Carbon Ecosystems in Hong Kong and the Greater Bay Area: Carbon Sequestration Capacity, Biogeochemical and Microbial Processes, and Control Mechanisms
Project Coordinator: Prof Hongbin Liu (HKUST)
Abstract
There is a global consensus that climate change, which threatens global ecosystems, human health, and economic development, is caused by greenhouse gas emissions, with anthropogenic CO2 arising from fossil fuel combustion being the largest contributor. To safeguard our planet and ensure its livability for future generations, governments worldwide have set goals to reduce greenhouse gas emissions over the coming decades. While measures such as using renewable energy instead of fossil fuels and improving energy efficiency in buildings, transportation, and waste treatment – Hong Kong’s three main energy-consuming sectors – can significantly reduce greenhouse gas emissions, it is not possible to achieve net zero carbon emissions without the help of Mother Nature.
Coastal blue carbon ecosystems are recognized for their great potential to enhance carbon sequestration and help achieve net zero carbon emissions. Our proposed project will study the past, present, and future of blue carbon ecosystems – mangroves and seagrasses – in Hong Kong, as well as the Greater Bay Area, focusing on their carbon sequestration capacity and the environmental threats that they face. We will link biogeochemical processes, microbial metabolic activities and pathways, and hydrodynamic forcing to achieve a unified and quantitative understanding of microbially driven biogeochemical processes in mangrove and seagrass sediments and reveal how these processes are influenced by physical forces such as tides and circulation patterns and anthropogenic impact including eutrophication and water pollution. We will also study the horizontal export of blue carbon to coastal seas and the potential greenhouse gas emissions from blue carbon ecosystems. Furthermore, we will develop pollution remediation strategies and multifunctional biochar-based microbial composites to optimize the restoration of coastal blue carbon ecosystems and enhance their carbon storage capacity.
The results of this study will shed light on the response of these valuable, yet vulnerable ecosystems to human impact and climate change. By the end of this research, we will provide a model predicting the contribution of blue carbon ecosystems in Hong Kong to the government’s goal of attaining carbon neutrality by 2050.
Topic 3: Establishing Hong Kong as the Leading Integrated Circuits, and Opto-electronics Innovation and Technology Hub in the Guangdong-Hong Kong-Macao Greater Bay Area
Project Title: Terahertz Photonic Chip for Integrated Sensing and Communications in the 6G Era
Project Coordinator: Prof Cheng Wang (CityU)
Abstract
The widespread adoption of 4/5G networks has profoundly reshaped the way people live and interact. Yet the drive for faster connectivity continues. Looking beyond 2030, 6G networks are expected to offer not only further increased data speeds, but also precise environmental sensing capability, enabling Integrated Sensing and Communications (ISAC). Both faster communication and high-resolution sensing necessitate shifting the wireless carrier to higher frequency bands, from under 5 GHz for 4G, to 5-30 GHz for 5G, and ultimately to millimeter- wave and terahertz (over 100 GHz) bands for 6G and beyond. However, upgrading current microwave systems to these higher frequencies presents substantial challenges, as devices become markedly more lossy and inefficient. As a result, new system architectures and device principles are urgently needed for 6G and beyond.
Photonic integrated circuits provide a promising solution to this, by performing the terahertz signal generation, reception, mixing, and processing tasks in chip-scale optical systems. This project aims to develop integrated terahertz photonic chips that meets the technological and economic demands of 6G, based on the thin-film lithium niobate (TFLN) platform. On the transmitter side, integrating TFLN modulators with uni-traveling carrier photodetectors, combined with co-packaged terahertz integrated circuits and antennas, will enable low-noise generation, amplification, and high-speed modulation of terahertz signals. On the receiver side, combining terahertz antenna arrays with modulators and signal processors will facilitate efficient signal reception, down conversion, and processing on a single TFLN chip. Advanced chiplet packaging, real-time spectral sensing, and adaptive spectrum management technologies will be investigated for intelligent terahertz photonic ISAC systems with enhanced situational awareness and multifunctionality. This endeavor will capitalize on the team's complementary expertise in photonics, terahertz, and integration technologies, alongside close collaborations with industrial partners and downstream users within the Greater Bay Area and around the globe.
Upon the successful completion of this project, we aim to deliver integrated terahertz photonic transmitter and receiver solutions ready for 6G applications. We will engage with our industrial collaborators to ensure their mass-producibility and commercial viability. The collaborative research, development and commercialization efforts will further establish Hong Kong as a leading hub for opto-electronics innovation and next-generation wireless technology in the Greater Bay Area.
Topic 3: Establishing Hong Kong as the Leading Integrated Circuits, and Opto-electronics Innovation and Technology Hub in the Guangdong-Hong Kong-Macao Greater Bay Area
Project Title: Technology-Driven RISC-V AI Architecture Innovations for Emerging Embodied Robotics Systems
Project Coordinator: Prof Yuan Xie (HKUST)
Abstract
This project seeks to transform the field of computer architecture by introducing RISC-V-based AI innovations for embodied robotics systems. As AI continues to advance, the demand for computational power is increasing exponentially, and traditional architectures such as x86 and ARM are nearing their performance limits. Moreover, the slowdown of technology scaling makes it imperative to explore new chip design approaches.
The research focuses on two key areas: RISC-V architecture and in-memory/near-memory computing. RISC-V, an open-source instruction set architecture (ISA), offers unparalleled flexibility for customizing AI workloads. In-memory and near-memory computing address the "memory wall" by reducing the amount of data movement between memory and processors, making them ideal for high-throughput AI tasks. These innovations promise enhanced performance, efficiency, and scalability for AI applications.
The project also leverages advanced packaging techniques such as hybrid bonding and chiplets, and explores new memory technologies like 3D vertical DRAM. These advancements overcome traditional scaling limitations and open up new possibilities for AI chip design.
By focusing on embodied robotics as the future of artificial general intelligence (AGI), the project addresses the computational and real-time processing challenges faced by interactive robots. An algorithm-architecture co-design approach will optimize AI systems to meet the specific needs of embodied robotics.
The deliverables include a RISC-V AI architectural simulator, design methodologies, open-source intellectual property (IP) cores, and chip prototyping. These contributions will benefit the research community and drive progress in the AI and semiconductor industries.
This project involves extensive collaboration with industry and is set to establish Hong Kong as a global hub for integrated circuits and optoelectronics innovation, contributing to the development of the next generation of AI systems capable of operating in complex environments.