Enhancing diagnostic capability of CSF dynamics in normal pressure hydrocephalus through advanced analysis of intracranial pressure waveforms and PC MRI Images

Summary

Normal Pressure Hydrocephalus (NPH) is a disease associated with disturbance in the cerebrospinal fluid (CSF) circulation. Uniquely, it is among the limited dementia types that can potentially be reversed via surgical intervention, specifically through shunt insertion (https://revertproject.org).

However, the effectiveness of this treatment is contingent upon precise and accurate diagnosis. Unfortunately, NPH is currently still under-appreciated and under-diagnosed, exacerbating The Brain Physics Lab at Cambridge has paved the way in innovating diagnostic techniques for NPH, leveraging the CSF infusion test. 

Project aims

This approach evaluates the intracranial pressure response upon administering artificial CSF at a constant rate, subsequently estimating parameters of a CSF circulation model. 

In addition, in collaboration with our colleagues at the University of Picardie Jules Verne, Amiens, France, we have implemented phase contrast MRI imaging for NPH patients to provide us with pulse morphology of cerebral fluids flows, complementing the pressure information obtained from the infusion tests. 

Metrics obtained from these methods offer critical insights for neurosurgeons contemplating the insertion or revision of a CSF shunt device. However, the predictive efficacy of these parameters, especially their negative predictive power, demands improvement. 

This project endeavours to augment the diagnostic methodology for NPH by development of more accurate and adaptive model identification methodologies to represent better the dynamics of CSF circulation, by exploration of models of brain compliance based on combination of the pressure (Infusion tests) and flow (PC MRI imaging) recordings from the brain, and by application of machine learning methods, trained on over 5000 infusion tests recordings, and rapidly growing number of pseudo simultaneous PC MRI images obtained, to build probabilistic models for waveform patterns classifications. 

Improving the accuracy of NPH diagnosis is imperative not only for better patient outcomes but also for minimizing unnecessary surgical interventions. 

A refined model that accurately predicts the need and potential success of shunt insertions may revolutionize the way we approach NPH, turning the tide in the battle against this reversible form of dementia.

The candidate will also work collaboratively with clinicians to validate findings and adjust models accordingly. 

Contact details

Dr Peter Smielewski - ps10011@cam.ac.uk

Opportunities

This project is open to applicants who want to do a:

• PhD
• MPhil

This project would suit graduates in Biomedical Engineering, Physics, Mathematics, Signal Processing, Electrical Engineering, and Computer Science, as well as mathematically minded medical graduates.