1 Postdoc and 2 PhD positions available (1.0FTE)

Information on the application procedure.

Big Question 1

Scientific summary
The big question this project addresses is how to use computational modelling to link levels of description, from neurons to cognition and behaviour, in understanding the language system. We focus on the mental lexicon and aim to characterise its structure in a way that is precise and meaningful in neurobiological and psycholinguistic terms. Our overarching goal is to devise causal/explanatory models of the mental lexicon that can explain neural and behavioural data. This will significantly deepen our understanding of the mental lexicon, lexical access, and lexical acquisition.

PhD Position - BQ1-1

Learning and Adaptation in Neurobiological Models of Language Processing

You will participate in a team effort to develop a computational neurobiology of language through the simulation of recurrent networks of spiking neurons for language processing.
The aim of the project is to understand the role that plasticity principles on different time scales might play in shaping processing memory for sentence comprehension. These principles include, for example, neuronal adaptation, short-term synaptic plasticity, inhibitory plasticity, long-term excitatory potentiation and depression, homeostasis and consolidation.
Of particular interest will be the interaction of plasticity with structural network features, for example, the connectivity structure in cortical microcircuits and spatial structure in the dendritic tree of neurons.
Computational models developed in the PhD project will be integrated into a larger architecture that includes a model of the long-term storage and retrieval of words. You will also investigate how these components interact during language processing.

You should have a Research Master’s degree (or equivalent) in neuroscience, physics, mathematics, engineering, or other relevant fields of study.
A high level of written and spoken English and excellent programming skills (e.g., Matlab, Python, Julia) are required. A strong background in computational neuroscience and computability theory is highly valued. You should be familiar with dynamical systems theory and differential calculus. You should also have active interest in language and psycholinguistics. Affinity with cognitive neuroscience and fundamental issues in theoretical linguistics is desirable. Previous experience with techniques for neural data analysis is desirable, but not mandatory.

Applications from excellent candidates with a less than ideal profile will also be considered.

Embedding and Terms of employment
This position will be held at the Max Planck Institute for Psycholinguistics Nijmegen; Neurobiology of Language Department.

  • Employment: 1,0 fte;
  • Starting salary will be based on the German TVöD (Tarifvertrag für den Öffentlichen Dienst). The amount will correspond with 50% of the amount in salary group E13, level 1, combined with a monthly 'Gewinnungszulage' (supplement); together approximately EUR 2.315 gross per month. From the second year the salary will be approximately EUR 2.565,- a month (50% of E13/level 2 + 'Gewinnungszulage').
  • In addition to the salary: an 8% holiday allowance
  • The Max Planck Institute involved has a number of regulations that make it possible for employees to create a good work-life balance.

Contact information
Dr. Karl Magnus Petersson and Dr. Hartmut Fitz

Big Question 4

Scientific summary
We aim to characterise variation in language processing and learning skills and to determine how these variations relate to those in the underlying biology of individual participants. The project has two strands: Strand A focuses on language processing skills in young adults, and Strand B on language learning skills in children and adults.
Strand A will develop a comprehensive battery of language tasks targeting sound, meaning, and grammatical processing of words and longer utterances during speaking and listening. In addition, tasks will be selected or developed assessing general cognitive skills that are likely to affect performance in language tasks. The battery will be normed on a demographically representative sample of 1000 young adults (aged 18-30 years). Strand B uses variability in learning ability to investigate why second-language acquisition can become harder in adulthood. It will consist of two sub-projects, one on grammar learning and one on word learning. In each sub-project, a large number of child, adolescent and adult Dutch participants (aged 8-30 years) will be tested using behavioural and neuroimaging techniques.

Postdoc Position - BQ4-1

In order to advance understanding of the relationships between quantitative imaging-derived readouts of brain structure, function and connectivity on the one hand and human cognition, behavioural readouts and developmental changes therein on the other hand, we propose to build ‘cognitive charts of language ability’. Our aim is to develop a comprehensive set of principled analysis techniques that are sensitive to subtle variations within and between functional brain networks, while incorporating rich multi-modal data and descriptions of subject-specific behavioural information. You will extend previous work in the ‘Toolkit’ work package of the consortium (normative modelling [Marquand et al. 2016]) to map inter-individual variation in performance on cognitive tasks to variability in underlying aspects of brain structure, function and connectivity in language-related networks as well as to other quantitative biological measures such as genome-wide genotype data. To do so, you will develop a set of statistical models that encode brain-behaviour mappings linking the different domains of language function to underlying neurobiological circuitry. Crucially, this position will extend previous normative models towards multi-variate characterisations that will take the dependence (covariance) structure of different cognitive/behavioural measures into account. The ideal candidate is expected to take the lead in regards to the methodological aspects of this project, addressing specifically the challenges related to the functional neuroimaging integration with behavioural and demographic information, incl. tool development, data analysis and manuscript writing.

The following requirements are mandatory:

  • PhD degree in Computer Science, AI, Statistics, Applied Mathematics Information Engineering, Neuroscience (with a strong analytical background), or a related field.
  • Excellent track record in machine learning or data science
  • Experience with functional neuroimaging in human participants and with neuroimaging data analysis techniques and software (i.e. SPM, FSL or similar);
  • Ability to work independently and to collaborate with other team members;
  • Excellent organizational and social skills;
  • Established publication record attesting the above-mentioned requirements;
  • Good English proficiency (especially scientific writing and oral communication).
  • Extensive experience in scientific computing, including numerical methods, large-scale simulations, and data-science approaches. Programming experience especially in languages such as Matlab and Python is required.

The following requirements are desirable:

  • Experience with multivariate data analysis approaches as implemented in established neuroimaging data analysis software;
  • Availability to start as soon as possible

Applications from excellent candidates with a less than ideal profile will also be considered.

Embedding and Terms of employment
This position will be held at the Donders Institute, Centre for Medical Neuroscience, Radboud UMC, Nijmegen, The Netherlands.

  • Employment: 1.0 FTE;
  • Scale 10: max € 58838 gross per year at full employment
    (36h per week, incl. 8% vacation bonus and 8.3% end of year payments);
  • you will be appointed for an initial period of 12 months, after which your performance will be evaluated. If the evaluation is positive, the contract will be extended by 24 months;
  • the Collective Labour Agreement (CAO) of Dutch Medical Centres is applicable to this position.

Contact Information
Prof. Christian F. Beckmann

PhD Position - BQ4-2

The development of a fundamentally new approach to understanding variability in clinical populations on the basis of biological measures derived from brain imaging will enable us to use data from large cohorts to learn a normative distribution that characterises population variation, without needing to make strong assumptions about subgroupings. In this project we will provide efficient tools and techniques to statistically quantify the nature of the underlying abnormalities at the level of individual subjects for very high-dimensional imaging phenotypes (voxel-wise maps or surfaces). This will scale normative models to whole-brain neuroimaging data while modelling their complex correlation structure. Using approaches from spatial statistics we will develop techniques for modelling spatially distributed imaging phenotypes. The resulting quantification of the spatial pattern will enable us to use the diagnostic labels as predictor variables for map-wise imaging results. This in turn will be used to develop automated approaches for cohort stratification independent of clinical labels beyond simple data-driven clustering approaches. We will make use of newly acquired data in BQ4 and fuse this with several already available, high quality datasets, incl. the Human Connectome Project data and several other data sets. Together we will have access to currently over 12000 datasets obtained in both healthy controls as well as clinical populations. Importantly, using local and remote data sources we will be able to include rich clinical datasets that provide a wealth of additional behavioural and clinical information. The analysis of these datasets will provide important benchmarks for the neuroimaging community as these datasets are being widely adopted for hypothesis generation and validation purposes.

You should have a Research Master’s degree (or equivalent) in Cognitive Neuroscience, Computer Science, Artificial Intelligence, Statistics or in any other related field of study.

Applicants should be able to demonstrate a strong academic track record and should have experience in analysing diffusion MRI/fMRI data. Experience with the usage of one or more of the common neuroimaging data analysis platforms (e.g. FSL, SPM, FreeSurfer) and programming skills in MATLAB, Python, R and/or C++ are highly desirable.

Applications from excellent candidates with a less than ideal profile will also be considered.

Embedding and Terms of employment
This position will be held at the Donders Institute, Centre for Medical Neuroscience, RadboudUMC, Nijmegen, Netherlands.

  • Employment: 1.0 FTE;
  • Scale 10A: max € 40814 gross per year at full employment
    (36h per week, incl. 8% vacation bonus and 8.3% end of year payments);
  • you will be appointed for a period of 48months;
  • the Collective Labour Agreement (CAO) of Dutch University Medical Centres is applicable to this position;

Contact information
Prof. Christian F. Beckmann and Prof. Jan Buitelaar