Postdoc in deep learning

Linköping University, Sweden
Location: 
Sweden
Job Type: 
Full Time
Closing Date: 
Monday, October 1, 2018

Description

In the EU-project IMPACT (which is 3 years long), LiU will use magnetic resonance imaging (MRI) to scan patients with brain tumours. The goal of the project is to develop new algorithms for detecting and segmenting (small) brain tumours, such that they can be treated before becoming large tumours. A specific goal is to develop algorithms, for example 3D convolutional neural networks (CNNs), that can take advantage of the fact that several types of 3D/4D data will be collected for every patient (structural MRI, diffusion MRI and possibly functional MRI). The project will be carried out in collaboration with the Swedish companies SyntheticMR, Inovia and Elekta.

The postdoc may also be involved in other research projects where medical image processing and machine learning are combined. One project is about diagnosis of brain diseases using multi-modal MRI data, and another project is about using CT data to automatically detect vertebra compression.

Qualifications

The position requires a doctorate or an equivalent degree from a foreign university. The doctorate shall have been obtained no longer than three years before the expiration date of the application.

The applicant is expected to have a PhD in (medical) image processing, computer science or machine learning. Experience of using CNNs and deep learning tools such as TensorFlow, Caffe and Keras is especially desirable. Experience of programming graphics cards using CUDA and OpenCL is desirable. Experience of GANs (generative adversarial networks) is desirable. Teaching experience is also a merit. The applicant should be able to work independently, be curious and have excellent collaboration skills, as the IMPACT project involves three Swedish companies.

Contact Anders Eklund for additional information, https://liu.se/en/employee/andek67

Apply through LiU, https://liu.se/en/work-at-liu/vacancies?rmpage=job&rmjob=9259&rmlang=UK