Hypoxic-ischaemic encephalopathy (HIE) affects babies' brains during the childbirth due to shortages of oxygen. Using computer vision and machine learning techniques, HIE disease is diagnosed much earlier than two years which is the current normal practice in hospitals. As a result of HIE early detection, then early interventions can be applied to improve the babies health.

Neonatal hypoxic-ischaemic encephalopathy (HIE) is a consequence of perinatal asphyxia and is a significant cause of perinatal death and neurodevelopmental impairments later in life. Approximately 0.2% of infants in high income countries suffer from HIE, and there is a mortality rate of 15%–25%. HIE carries a high risk for neuromotor, cognitive, and behavioural difficulties, epilepsy, visual and hearing impairment in survivors. Early diagnosis of the injury location and extent is important for counselling and identification of those who may benefit from early intervention. A range of techniques are used for diagnostic evaluation, including magnetic resonance imaging (MRI) with T1 In this document, we are proposing to develop a method for HIE detection by analysing Susceptibility Weighted Images (SWIs). In the current clinical practice, the detection of HIE is performed 24 months after the birth by evaluating the child’s behaviour. In this document, we propose a framework to detect HIE after the birth for infants who may have suffered asphyxia during the birth by analysing the MRI images of their brains after the birth. The early detection of HIE in new-borns helps health carers to intervene early to improve the prognosis of HIE and reduce the occurrence of sequelae. We also propose methods to find the regions where the brain injuries have occurred to enable the health-carers to predict what impacts these injuries might have on the patients’ behaviours and therefore to provide a more targeted intervention to aim for a better outcome at the age of 24 months.

Co-supervisor: Prof Bridget Vollmer  ( https://www.southampton.ac.uk/people/5x7s26/professor-brigitte-vollmer )

Entry requirements

You must have a UK 2:1 honours degree, or its international equivalent

You need to have a strong background and skills in:

• machine learning
• computer vision
• Python programming
• Deep Learning

How to apply Hypoxic-ischaemic encephalopathy early detection by using image analysis of MRI susceptibility weighted imaging | University of Southampton

You need to:

• choose programme type (Research), 2025/26, Faculty of Engineering and Physical Sciences
• select Full time or Part time
• choose the relevant PhD in Computer Science
• add name of the supervisor Dr Sasan Mahmoodi in section 2 of the application form

Applications should include:

• personal statement
• your CV (resumé)
• 2 academic references
• degree transcripts to date