Research Associate: Medical Imaging AI - CAPTURE-PH RAIDA

University of Sheffield; Sheffield; Hybrid; GBP 42,000 - 52,000; Full time

Job summary

The University of Sheffield invites applications for a two-year Grade 7 Research Associate position in CAPTURE-PH to translate and refine RAIDA cardiothoracic CT AI across large, clinically characterised pulmonary hypertension cohorts. You will audit, retrain and technically harden existing RAIDA CT assets so they operate robustly across ASPIRE and PHINDER, tackling scanner heterogeneity and data quality.

Applicants should have a PhD or equivalent in computer science, machine learning or medical

Qualifications

• PhD or equivalent in computer science, ML, medical imaging or related field.
• Strong Python/PyTorch skills and 3D medical imaging experience.
• Experience with robust model validation and reproducible research software.
• Familiarity with nnU-Net/MONAI and explainable AI is advantageous.

Responsibilities

• Audit, refine, retrain and harden RAIDA CT assets for cross-centre robustness.
• Develop and evaluate segmentation, feature-extraction and prediction pipelines for biomarkers.
• Apply domain adaptation, semi-supervised learning and human-in-the-loop refinement.
• Link imaging outputs with clinical data to support CAPTURE-PH questions.
• Establish reproducible multi-centre validation workflows.

Skills

Python

PyTorch

3D medical imaging

Model validation

Domain adaptation

Semi-supervised learning

Explainable AI

Reproducible research

Education

PhD or equivalent

Tools

nnU-Net

MONAI

Job description

Are you a computer scientist or machine-learning researcher interested in making medical-imaging AI work reliably beyond the dataset on which it was developed?

We have an exciting two-year Grade 7 Research Associate opportunity within CAPTURE-PH, focused on translating and refining existing RAIDA cardiothoracic CT AI across large, clinically characterised pulmonary hypertension cohorts.

RAIDA is our existing cardiothoracic imaging-AI programme, bringing together automated analysis of the heart, pulmonary vasculature and lung parenchyma with explainable outputs and expert review. You will not be starting from a blank sheet: your core responsibility will be to audit, refine, retrain and technically harden existing RAIDA CT assets so that they operate robustly across ASPIRE and PHINDER, with extension to other approved CAPTURE-PH cohorts. You will tackle scanner and protocol heterogeneity, data-quality issues and model failure modes, and establish reproducible multi-centre validation workflows.

You will develop and evaluate segmentation, feature-extraction and prediction pipelines for lung, cardiac and pulmonary vascular biomarkers, using expert-reviewed imaging and clinically rich reference data. Where appropriate, you will use approaches such as domain adaptation, semi-supervised learning and human-in-the-loop refinement to improve generalisability. You will link imaging outputs with right-heart catheterisation, lung function, walk testing, MRI and outcomes to support CAPTURE-PH questions around diagnosis, disease severity, prognosis and treatment-responsive phenotypes.

We are looking for someone with a PhD or equivalent relevant experience in computer science, machine learning, medical imaging, biomedical engineering or a related discipline. You will have strong Python/PyTorch skills, experience with 3D medical imaging and robust model validation, and the ability to build reproducible research software.

Experience of nnU-Net/MONAI, explainable AI, multi-centre model adaptation, thoracic CT or clinical AI validation would be an advantage.

The post is fixed-term for 24 months and full-time. We are committed to exploring flexible working opportunities which benefit the individual and University, subject to the requirements of secure research data and computing environments.

The University of Sheffield offers a generous benefits package, including annual leave, pension provision, flexible-working opportunities and support for professional development.

We build teams of people from different heritages and lifestyles from across the world, whose talent and contributions complement each other to greatest effect. We believe diversity in all its forms delivers greater impact through research, teaching and student experience.