Programme(s) to which this project applies:
|☑ MPhil/PhD||☒ MRes[Med]||☒ URIS|
Osteoporotic fractures incur significant patient morbidity and healthcare costs. Predicting which patients are at risk of fracture is of paramount importance as effective drug treatment exists in reducing fracture risk, but not without side effects if used indiscriminately. Unfortunately, current models developed from conventional statistical methods fall short in terms of accuracy, with reported sensitivity of 62.3% and positive predictive value of 10.3%. Fracture is a complex interplay of myriad patient and environmental factors. In Hong Kong, public hospitals manage virtually all patients with medical emergencies, and capture territory-wide data from electronic health records across 43 hospitals. This offers an opportunity of harnessing information from a dataset of over 7 million, harbouring anonymised real-life patient information including medical diagnoses and comorbidities, number of hospital admissions, surgical operations, duration of in-hospital stay, number of out-patient appointments, drug prescriptions, investigations performed, etc. Different machine learning methods could be employed to explore complex non-linear relationships between patient characteristics and their subsequent fracture risk, in hopes of building a fracture prediction model of superior accuracy than best existing methods. Such a model would allow timely intervention, with significant impact to patients and on the public health level.
Dr JSH Wong, Department of Orthopaedics and Traumatology
Janus is a Clinical Assistant Professor at the Department of Orthopaedics & Traumatology at the University of Hong Kong. A doctor by training and member of the Royal College of Surgeons of Edinburgh, Janus treats and performs orthopaedic operations on patients on a daily basis. He has an interest in applying machine learning methods to address challenges in an evolving clinical landscape. He has received recognition and published in international journals on the use of ‘big data’ to improve patient care and clinical practice.
For more information or to express interest for this project, please email the supervisor or the specified contact point in the project description. Interested candidates are advised to enclose with your email:
Information on the research programme, funding support and admission documentations could be referenced online at the Research Postgraduate Admissions website.
General admission enquiries should be directed to email@example.com.