Postdoc: ML for Psychosis Prediction & Brain Connectivity

Jens Gilles; Greater London; Ibrido; GBP 36.000 - 48.000; Tempo pieno

Descrizione del lavoro

Jens Gilles in London invites applications for a 2-year postdoctoral research post on the UKRI-funded PROSPECT project. The role aims to develop machine learning methods to predict psychosis outcomes from MRI-derived brain connectivity, with potential extension to 4 years.

The position sits at the intersection of ML, neuroimaging and psychiatry, handling rich MRI datasets, data curation, and network construction using the Morphometric Inverse Divergence (MIND) approach.

Competenze

• PhD (or near completion) in a relevant subject area (CS, Engineering, Physics, Mathematics, Psychiatry, Psychology or Neuroscience).
• Experience working with brain MRI data.
• Excellent computational skills with ML libraries in Python (scikit-learn, PyTorch, TensorFlow).
• Experience writing up research results for publication.
• Ability to communicate technical results to researchers at seminars/conferences.
• Strong organisational and time management skills.
• Ability to work independently and creatively, and to collaborate in a team.
• Ongoing professional development in the specialist area.

Mansioni

• Develop innovative machine learning approaches to predict longitudinal symptom changes.
• Work with rich brain MRI datasets, curate and process data, derive functional and structural brain networks.
• Apply Morphometric Inverse Divergence (MIND) for estimating structural networks from T1-weighted images.
• Explore uncertainty-aware ML to improve prediction accuracy on subsets of patients.
• Test novel ways to define longitudinal outcomes and combine functional and structural networks.
• Optionally relate brain imaging data to speech and language data if appropriate.

Formazione

PhD (or near completion) in a relevant subject

Strumenti

Python (scikit-learn, PyTorch, TensorFlow)

Brain MRI data processing

Descrizione del lavoro

Jens Gilles in London invites applications for a 2-year postdoctoral research post on the UKRI-funded PROSPECT project. The role aims to develop machine learning methods to predict psychosis outcomes from MRI-derived brain connectivity, with potential extension to 4 years.

The position sits at the intersection of ML, neuroimaging and psychiatry, handling rich MRI datasets, data curation, and network construction using the Morphometric Inverse Divergence (MIND) approach.