Research Associate in Multimodal Agentic AI for Cardiology

University of Manchester; Manchester; Hybrid; GBP 40,000 - 50,000; Full time

Benefits offered by this job

Fantastic market leading Pension

Excellent employee health and Wellbein

Exceptional starting annual leave Ent.

Additional paid closure over Christmas

Discounts at major retailers

Job summary

The Division of Informatics, Imaging and Data Sciences at the University of Manchester invites applications for a Research Associate to develop multimodal agentic AI for clinical decision support in cardiology. You will integrate cardiac imaging, ECGs and clinical data to form AI systems that aid clinician review.

You will curate cardiovascular datasets, design AI agents that gather evidence and reason with analytical tools, and evaluate accuracy, reliability and clinical relevance.

Qualifications

• PhD in a related discipline or nearing completion.
• Strong Python programming skills.
• Experience in deep learning and AI methods.
• Ability to design and critically evaluate AI systems.
• Experience with multimodal learning or medical imaging is advantageous.

Responsibilities

• Curate multimodal cardiovascular datasets.
• Develop methods to integrate imaging and clinical data.
• Design AI agents for evidence gathering and clinical reasoning.
• Publish findings and support grant applications.

Skills

Python

Deep learning

Multimodal learning

AI agents

Medical imaging

Education

PhD in CS/AI/Medical imaging/ Biomedical Engineering / Applied Mathematics

Job description

The Division of Informatics, Imaging and Data Sciences within the Faculty of Biology, Medicine and Health invites applications for a Research Associate to develop multimodal agentic AI for clinical decision support in cardiology. The project will integrate cardiac imaging, ECGs, clinical reports and other patient information to develop AI systems that interpret clinical evidence, identify missing or conflicting information, and recommend appropriate next investigations for clinician review.

In this role, you will curate multimodal cardiovascular datasets, develop methods to integrate imaging and clinical information, and design AI agents that gather evidence, use analytical tools and support clinical reasoning. Working closely with cardiologists, you will evaluate these systems for accuracy, reliability, interpretability and clinical relevance. You will also contribute to research publications, present findings at conferences and support the development of future funding applications.

The successful candidate will work under the supervision of Dr Jinming Duan, collaborating with colleagues in the BHF Manchester Centre of Research Excellence and clinical teams at Wythenshawe Hospital, part of Manchester University NHS Foundation Trust.

You should hold, or be nearing completion of, a PhD in computer science, artificial intelligence, medical image analysis, biomedical engineering, applied mathematics or a related discipline. You will have strong Python programming skills, practical experience in deep learning, and the ability to design and critically evaluate AI methods. Experience with multimodal learning, large language models, AI agents or medical imaging would be advantageous. Clear communication, effective collaboration and a commitment to rigorous, reproducible research are essential.

What you will get in return:

• Fantastic market leading Pension scheme
• Excellent employee health and wellbeing services including an Employee Assistance Programme
• Exceptional starting annual leave entitlement, plus bank holidays
• Additional paid closure over the Christmas period
• Local and national discounts at a range of major retailers

As an equal opportunities employer we welcome applicants from all sections of the community regardless of age, sex, gender (or gender identity), ethnicity, disability, sexual orientation and transgender status. All appointments are made on merit.

Our University is positive about flexible working – you can find out more here

Hybrid working arrangements may be considered.