Research

Novel medical image analysis

Quantitative analysis of MRI, PET and CT data to generate useful metabolic (e.g. pH) and hemodynamic (e.g. cerebral blood flow) information about the body for clinical applications such as stroke and cancer diagnosis.

Machine learning in clinical and big data applications

Using artificial intelligence and signal processing techniques on very large quantities of data that are routinely collected by healthcare devices and cloud systems to provide automated, semi-automated and decision support systems for:

• patient monitoring in clinical applications
• data file classification and user behavior prediction

Sound recovery via negligible object vibration in videos

The majority of the installed CCTVs usually just record the imagery but not sound, especially the outdoor cameras because these cameras are normally fixed at a distance either high-up or far away from the recording scene, resulting in mainly random noise such as wind or inaudible sound being recorded.

The main aim of this research is to equip the outdoor surveillance system to have not only "eyes", but also "ears" to assist in scientific evidence collection during an investigation by extracting/recovering sound from negligible object vibrations caused by the sound waves via imagery.

Physiological and mathematical modelling

Mathematical modelling of cilia using computational fluid dynamic (CFD) tools such as CFD-ACE, to study the protein-structure-fluid interaction in a healthy and diseased state in order to understand the diseases caused by defective and dysfunctional cilia.