A Machine Learning Platform for Personalised Surgical Decision-Making in Hydrocephalus

Summary

Motivation

Hydrocephalus is a debilitating neurological condition with diverse aetiologies, including post-haemorrhagic hydrocephalus in neonates (PHH/IVH), idiopathic Normal Pressure Hydrocephalus (iNPH) in the elderly, and secondary cases following subarachnoid or intracerebral haemorrhage (SAH/ICH).

Current management is hampered by significant uncertainty in predicting which patients will benefit from surgical intervention (shunting), leading to variable outcomes, unnecessary procedures, and avoidable complications.

While recent research, demonstrates the promising application of Machine Learning (ML) for predicting outcomes like shunt responsiveness, these models are typically single-centre, focused on a single hydrocephalus phenotype, and lack integration into clinical workflows.

They often fail to provide clinicians with the calibrated probabilities and uncertainty measures needed for confident, personalised decision-making.

This project aims to bridge this critical gap between algorithmic promise and clinical utility.

Objective

The primary objective is to develop, calibrate, and externally validate a multimodal machine learning platform for clinically actionable outcome prediction across the major hydrocephalus phenotypes.

Specific objectives are:

1. To curate a large, retrospective dataset comprising clinical, radiological (including CT/MRI features and volumetric analyses), and laboratory data from pediatric (PHH/IVH), iNPH, and post-SAH/ICH patient cohorts. This has been secured.

2. To develop and train phenotype-specific ML models to predict key clinical outcomes, such as shunt responsiveness, functional improvement (e.g., on the modified Rankin Scale), and complication risk.

3. To rigorously validate these models externally on unseen data from partner institutions and benchmark their performance against existing standard clinical scoring systems (e.g., CHESS, SDASH, mCPPRH).

4. To implement critical translational elements, including pre-specified clinical decision thresholds, decision-curve analysis to evaluate clinical net benefit, uncertainty quantification for each prediction, and fairness reporting to audit for bias across demographic subgroups.

Outputs

The project will generate a suite of tangible outputs with significant academic and clinical impact:

• A Validated ML Software Tool: A prototype software platform that integrates with hospital systems to provide clinicians with interpretable risk predictions for individual patients.
• Open-Source Code and Model Weights: To ensure reproducibility and foster further research, the code and non-proprietary model architectures will be made publicly available on a platform like GitHub.
• Clinical Decision Support Framework: A comprehensive framework for implementing ML-based decision support in neurosurgery, including guidelines on interpreting model output, uncertainty, and fairness metrics.
• A Large, Annotated Multi-Centre Dataset: A foundational dataset that will serve as a valuable resource for future research in neuroinformatics.

AccessNI clearance required

Please note, the successful candidate will be required to obtain AccessNI clearance prior to registration due to the nature of the project.

Essential criteria

Applicants should hold, or expect to obtain, a First or Upper Second Class Honours Degree in a subject relevant to the proposed area of study.

We may also consider applications from those who hold equivalent qualifications, for example, a Lower Second Class Honours Degree plus a Master’s Degree with Distinction.

In exceptional circumstances, the University may consider a portfolio of evidence from applicants who have appropriate professional experience which is equivalent to the learning outcomes of an Honours degree in lieu of academic qualifications.

• Sound understanding of subject area as evidenced by a comprehensive research proposal
• A comprehensive and articulate personal statement

Desirable Criteria

If the University receives a large number of applicants for the project, the following desirable criteria may be applied to shortlist applicants for interview.

• First Class Honours (1st) Degree
• Completion of Masters at a level equivalent to commendation or distinction at Ulster
• Practice-based research experience and/or dissemination
• Experience using research methods or other approaches relevant to the subject domain
• Work experience relevant to the proposed project
• Publications record appropriate to career stage

Equal Opportunities

The University is an equal opportunities employer and welcomes applicants from all sections of the community, particularly from those with disabilities.

Appointment will be made on merit.

Funding and eligibility

NOTE - This is a self-funded research project and applicants will be required to provide evidence of funds to support their tuition fees and living expenses.

Applicants should hold, or expect to obtain, a First or Upper Second Class Honours Degree in a subject relevant to the proposed area of study.