Senior Multimodal AI Scientist – Computational Radiology Full-time

Senior Multimodal AI Scientist – Computational Radiology

At AstraZeneca, we put patients first and strive to meet their unmet needs worldwide. Working here means being entrepreneurial, thinking big and working together to make the impossible a reality. If you are swift to action, confident to lead, willing to collaborate, and curious about what science can do, then you're our kind of person.

We are seeking an AI and machine learning scientist to develop computational biomarkers and predictive models from multimodal biomedical data, including radiology imaging, clinical, molecular, and other patient-level data, for our Computational Radiology team within Biomarker Sciences & Technologies (BST) group. Our team plays a crucial role in supporting AstraZeneca's early oncology and late development strategy for an innovative pipeline that includes Antibody-Drug Conjugates (ADCs), Radio-conjugates, T-cell engagers, CAR-T therapies, bispecific antibodies, and small molecules.

In this role, based in Boston, MA, you will collaborate with a diverse team of radiologists, imaging scientists, radiation physicists, translational scientists, biologists, and oncologists. This unique opportunity allows you to contribute to the development of new biomarkers, enabling indication selection, early assessment of biological activity, and optimal patient stratification. Your efforts will significantly enhance the probability of success for AstraZeneca's oncology pipeline.

The “Sr. Multimodal AI Scientist – Computational Radiology” will work to leverage foundational and cutting-edge techniques to drive the development of computational biomarkers and advanced predictive models by integrating radiology imaging with clinical, molecular, pathology, and other biomedical data source in combination with business domain knowledge, to develop and apply advanced modelling and simulation algorithms (e.g. deep learning, foundational models, traditional Machine learning including classification, regression, clustering, graph theory, Monte-Carlo sampling, and more) to generate business and scientific insights. The role will work within defined project scope and solutions aligned to established governance frameworks and policies.

Responsibilities:

• Lead the design, development, and validation of computational pipelines that generate robust biomarkers and predictive models from multimodal biomedical data, including imaging, clinical, molecular, and real-world datasets.
• Develop machine learning and statistical modeling approaches that identify patient subgroups, predict outcomes, and generate clinically actionable insights from high-dimensional multimodal datasets.
• Develop, implement, and support modeling solutions that interrogate complex, multimodal datasets to generate scientific and business insights, applying modern machine learning, statistical learning, representation learning, foundation models, causal inference, and related computational approaches where appropriate.
• Design and implement multimodal analytical frameworks that integrate imaging data with clinical, molecular, and other non-imaging data sources to support patient stratification and endpoint prediction.
• Researching and developing predictive and explainable computational methods to guide decision-making within project parameters and established approaches.
• Present or publish findings for conferences and in peer reviewed journals.
• Builds effective relationships with established range of stakeholders to ensure utilization and value of information resources and services. Clearly and objectively communicate results, as well as their associated model assumptions, uncertainties and limitations within agreed frameworks.
• Develop, maintain, and apply ongoing knowledge and awareness in trends, standard methodology and new developments in analytics and data science
• Implement good working practices to ensure that computational radiology work is delivered to robust quality standards and aligned to defined governance frameworks and policies
• Collaborates in a multidisciplinary environment with world leading clinicians, data scientists and statisticians, biological experts, clinical trial delivery teams, and IT professionals.

Qualifications:

• PhD in Computer Science, Statistics, Biostatistics, Machine Learning, Biomedical Engineering, Computational Biology, Bioinformatics, Applied Mathematics, Physics, or related quantitative discipline with 0-1 years of experience in the industry with a strong foundation in machine learning, statistical modeling, applied mathematics, computer vision, computational biology, bioinformatics, biomedical engineering, or related quantitative disciplines.
• Demonstrated experience building end-to-end ML pipelines including data preprocessing, model development, validation and performance assessment.
• Practical software development skills in standard data science tools: Python, R with demonstrable knowledge of good coding practices.
• Strong track record of publications in quality conferences and journals.
• Strong communication skills: ability to present compelling cases to collaborators and operate dynamically to identify solutions. Ability to work effectively within a team.

Desirable for the role:

• Prior experience in medical imaging, computational pathology, genomics, clinical data science, or other biomedical data domains is valued.
• Experience developing and applying machine learning approaches in multimodal analytical settings, integrating medical imaging with clinical, molecular, or other biomedical data sources.
• Expertise in statistical learning methods for high-dimensional data, including multiple testing correction and feature selection.