Three PhD positions: Child Health through Artificial intelligence Models and Predictions (CHAMP) (1.0 FTE)

At the department of Developmental Psychology at Utrecht University we are looking for highly motivated candidates interested in one of three PhD projects as part of the Child Health through Artificial intelligence Models and Predictions (CHAMP) project. The CHAMP project, awarded with a NGF AiNed Fellowship grant to Dr. Sonja de Zwarte, aims to develop and apply deep learning techniques to extract valuable information on early brain and social-emotional development from already acquired images and videos within the YOUth cohort study to predict mental health outcome in children.

The three PhD vacancies each have their own focus, but the PhD candidates will work closely together in an interdisciplinary team of AI-specialists, computer scientists, psychologists, neuroscientists and environmental epidemiologists on the CHAMP project, all supervised by Dr. Sonja de Zwarte.

• PhD1:
  

You will study early brain development based on analyzing 3D fetal ultrasound data. You will be co-designing and improving deep learning methods for automatic and rapid segmentation of the brain in fetal 3D ultrasound. In particular, you will be involved in improving our current convolutional neural network to quantify fetal brain structures from 3D ultrasounds in the YOUth cohort, applying the neural network to independent populations and clinical data sets, exploring unsupervised transfer learning methods and exploring the added value of multimodality imaging by combining fetal MRI and ultrasound data. You will become part of an existing collaboration between Utrecht University and the Eindhoven University of Technology (TU/e). Professor Hilleke Hulshoff Pol (Experimental Psychology, Utrecht University) and Dr. Ruud van Sloun (Electrical Engineering, TU/e) will be the promotors of the PhD trajectory. You will have regular interaction with the Signal Processing Systems group at the Electrical Engineering department at the TU/e.

• PhD2:
  

You will study early social-emotional development based on analyzing parent-child interaction videos. You will be co-designing and improving deep learning methods for automatic and rapid classification of parent-child interaction videos available in the YOUth cohort to quantify emotional face expression and early motor developmental features in young children. You will build upon existing work on child behaviour analysis under different activities, as well as the estimation of child affect during natural interactions. In addition, self-supervised feature extraction with contrastive learning techniques, including contrastive predictive coding and SimCLR, will be explored. Ultimately, you will aim to integrate longitudinal measurements from video material to follow individual development of young children over time. Professor Jaap Denissen (Developmental Psychology, Utrecht University) and Professor Albert Salah (Information and Computing Sciences, Utrecht University) will be the promotors of the PhD trajectory. You will have regular interaction with the Social and Affective Computing group at the Department of Information and Computing Sciences.

• PhD3:
  

You will study environmental factors (i.e., exposome) impacting early childhood development. You will develop prediction machine learning models based on the available imaging data and investigate how the exposome influences early brain and social-emotional development. The exposome concept represents the sum of all environmental drivers of health and disease, and, together with recent technological advances, has the potential to identify environmental contributors to health and disease in a manner complementary to the genome. For characterizing the exposome, you will e.g., link spatio-temporal characterizations of the social, food and physico-chemical environments with mental health outcome in children. Unraveling which environmental factors have positive or adverse effects on the developing child will aid in the design of effective prevention strategies. Professor Hilleke Hulshoff Pol (Experimental Psychology, Utrecht University), Dr. Odilia Laceulle (Developmental Psychology, Utrecht University), and Dr. Ulrike Gehring (Institute for Risk Assessment Sciences, Utrecht University) will be the (co-)promotors of the PhD trajectory. 

You will become skilled in developing hypotheses and analyzing large scale datasets, publishing in peer-reviewed scientific journals, and presenting your work at conferences. The position allows you to obtain an academic qualification (PhD). In addition, the CHAMP project offers opportunities to develop yourself in terms of interdisciplinary research, societal impact, outreach, and open science.

Your work will also include a 10% teaching task, such as the supervision of Bachelor’s and Master’s theses. Of course, you will be guided and supported by your supervisors and colleagues.

You will work in a collaborative, social, and dedicated team. We will guide you in your research and teaching tasks, which will help you develop your academic career. Each PhD candidate will benefit from courses from the Utrecht University graduate school.

We are looking for enthusiastic and collaborative team members who meet the following requirements:

• (Research) Master’s degree in the field of artificial intelligence, computer science, mathematics, statistics, psychology, neuroscience, epidemiology or equivalent programme;
• Strong interest and motivation to conduct interdisciplinary research at the intersection of childhood development and artificial intelligence;
• Strong affinity for and experience with programming (e.g., Python, PyTorch, TensorFlow, R, Matlab);
• Ideally: background knowledge of (developmental) psychology in terms of commonly used theories and methods, and the ability to create linkage between new methods and theoretical questions;
• Knowledge of and affinity with quantitative research and statistics and a high motivation to learn new methods;
• Good communication abilities and affinity with working in a multidisciplinary research team while maintaining autonomy;
• Strong proactive, collaborative, and creative working approach;
• High proficiency in written and spoken English;
• Interest and motivation to contribute to the teaching activities of the faculty.

There is a difference regarding the emphasis of AI for each position:

• PhD1 involves advanced 3D (medical) image processing.
• PhD2 involves advanced video processing, and both require a strong computational background and ample experience with machine learning methods.
• PhD3’s main focus will be on developing accurate prediction models based on repeated measurements over time.