Brain Tissue Pulsations: modelling and machine learning methods for the detection of raised intracranial pressure in adult intensive care.

Aims: This PhD project aims to develop criteria for the detection of raised Intracranial Pressure (ICP), based on non-invasive ultrasound measurements of brain tissue pulsations, for automated detection of raised ICP in patients with acquired brain injury.

Research Plan The brain is formed of soft neural tissue that is perfused with blood, surrounded by cerebrospinal fluid (CSF), and encased within the skull. The pulsating dynamics of brain tissue is determined by the neuroanatomy of the region in which the measurements take place, the compliance of brain tissue, and the dynamics of CSF, venous and arterial blood flow (Fig. 1(a)).
The first year of the PhD will involve a review of existing models of cerebral flows dynamics, and familiarisation with the modelling software (Matlab). In the second year, we will explore the hydro-mechanical components coupling cerebral flows to brain tissue pulsatility. Over the first two years, in parallel with the physical modelling task, we will also continue the exploration of the brain pulsatility features with a machine learning approach, which was initiated previously.2,3 In the third year, modelling and machine learning results, will be used to construct a classifier to detect raised ICP. The performance of the classifier will be validated against invasive ICP measurements from head trauma patients at Nottingham QMC, under the supervision of Professor Stuart Smith, who will provide clinical input to the study.
This PhD opportunity will be widely advertised amongst undergraduate and final year project students at the University of Leicester, and promising students will be personally approached to identify committed high quality candidates.