ML clinical prediction models that are trained on a specific patient population may not offer reliable prediction if the data distribution changes after the model is deployed. This points to the need for:

i) solid theoretical understanding of models' behaviour under data shift

and

ii) implementation of OOD detection models to determine to what extent the model is reliable.

The candidate will contribute to the solution to this problem by:

1. Investigating the desired properties for OOD detection models and improving on current techniques;
   
   
2. Carrying out experimentation to test which OOD detection methods work in practice on medical data.
   
   

The project will focus on Electronic Health Records from various medical sources. The goal of the project is to improve our understanding of the OOD detection problem and enable safe(r) deployment of medical AI tools by providing concrete and implementable solutions.

For examples of papers on the topic see here and here.

We are looking for a PhD candidate who is eager to contribute to the safer implementation of ML methods in healthcare settings.

You will investigate how to build a suitable OOD detection model that is meant to act as a failsafe layer, enabling the user of a ML model in healthcare to know to what degree the model in question is reliable. This exploration will address a number of methodological topics; a non-exhaustive list includes:

• Robustness under data shift.
  
  
• Density estimation.
  
  
• Interpretability and explainability for OOD detectors.
  
  
• Uncertainty estimation and model confidence.
  
  
• Counterfactual and/or adversarial methods for model robustness.
  
  
• Domain adaptation.
  
  

You will provide efficient and scalable implementations of your methods and will integrate them with popular open-source systems. Via your affiliation with the Medical Informatics department, you will have access to several large datasets comprising tens of thousands of patient records.

You will be integrated in an ongoing collaboration with industrial partners, with concrete possibilities for your research to influence the development of existing medical AI products. Finally, you will be integrated in the vibrant community of the Medical Informatics department, with the opportunity to interact with talented researchers and to contribute to top-rated education programs.

Prerequisites:

• Master's degree in computer science, machine learning, artificial intelligence, mathematics, data science, medical informatics, or related fields.
• Hands-on experience in Machine Learning.
• Strong programming skills in Python and machine learning libraries (e.g., PyTorch, Sklearn).
• Creative and independent mindset.
• Capable of working in multidisciplinary environments.
• Fluency in English.

Not required, but helpful:

• Experience with responsible AI (e.g., fairness, explainability, causal inference).
• Experience working with real-world data, e.g. having experience with data wrangling, analysis, visualizations, etc. Even better if the experience is with medical data.
• Experience with scientific publications.
• Contribution to open-source projects.
• Willingness to learn Dutch.
• Excited to work on a problem with societal impact.

• Plenty of room for your drive to shape tomorrow's healthcare.
• Working on large-scale and in-house research, with motivated colleagues from all corners of the world.
• You will start with a contract for one year (12 months) in accordance with the CAO UMC 2022-2023, with the possibility of extension for another three years (36 months).
• PhD students (Onderzoeker in Opleiding) are placed in scale 21, with a fulltime gross salary. The starting salary is € 2.631,- and increases to € 3.336,- in the fourth year.
• Besides a good basic salary, you will also receive 8.3% end-of-year bonus and 8% holiday allowance. Calculate your net salary here.
• Pension accrual with BeFrank, a modern, comprehensible and fairly priced pension.
• Excellent accessibility by public transport and reimbursement of a large part of your travel expenses. We also have sufficient parking spaces at the AMC location and a good bicycle scheme.

View all terms of employment at AMR Arbeidsvoorwaarden AMR - Amsterdam UMC

Amsterdam UMC

You will be appointed at the Medical Informatics department of the Amsterdam UMC at the University of Amsterdam.

You will be supervised by dr. Giovanni Cinà, prof. Ameen Abu-Hanna and prof. N de Keizer. Dr. Cinà is Assistant Professor in AI and works on the reliability of medical AI applications, and specifically on topics such as OOD detection, Explainable AI and Causal Inference. He holds a joint position with Pacmed and the Institute for Logic, Language and Computation. Prof. Abu-Hanna is a Principal Investigator in Methodology in Medical Informatics and has extensive experience in AI, ML, and prognostic modelling and evaluation. Prof. de Keizer is a Principal Investigator in Medical Informatics, focusing on the evaluation of quality of care and leading NICE, the national registry for Intensive Care evaluation.

Our team has on-going collaborations with colleagues at the IvI-UvA, the ILLC-UvA and the Amsterdam Business School, and you are encouraged to actively participate in these collaborations.