Three postdoctoral fellowships in mathematics of medical imaging

Work environment

KTH Royal Institute of Technology in Stockholm is the largest and oldest technical university in Sweden. No less than one-third of Sweden’s technical research and engineering education capacity at university level is provided by KTH. Education and research spans from natural sciences to all branches of engineering and includes architecture, industrial management and urban planning. There are a total of almost 11,500 first and second level students and almost 1,900 doctoral students. KTH has 4,900 employees.

The School of Engineering Sciences carries out a wide range of research at the international front line, from fundamental disciplines such as physics and mathematics, to engineering mechanics with applications such as aeronautics and vehicle engineering. We also offer University degree programs in Engineering Physics, Vehicle Engineering, and 'Open Entrance', as well as a number of International Masters Programmes.

The Department of Mathematics at KTH has an exceptionally good international reputation and is a leading department in northern Europe. Notable awards are the 2006 Abel prize to Lennart Carleson and the 2010 Fields medal to Stanislav Smirnov for work during his time at the department. Research being carried is very broad in scope, encompassing essentially all fields of mathematics, with many related sciences also represented. Among the particularly strong research areas one finds dynamical systems and stochastics, diverse aspects of geometry, combinatorics, biostochastics, actuarial & financial mathematics, and computational mathematics with numerical analysis and theoretical computer science. The department also hosts a center, Center for Industrial and Applied Mathematics, that serves as catalyst for the integration of mathematics and its applications in academia, industry, and society in general.

Program Description

The Department of Mathematics at KTH -- Royal Institute of Technology (KTH) Stockholm, Sweden would like to invite applications for three Postdoctoral Scholarships starting either in September 1, 2014 or January 1, 2015.

The positions are part of the five year program "Low complexity image reconstruction in medical imaging'' financed by the Swedish Foundation for Strategic Research for collaboration between KTH, Karolinska University Hospital (KUS), Karolinska Institutet (KI) and Elekta.

The mathematical research centers on penalizing image complexity in tomographic reconstruction by sparsity and/or entropy. Topics of interest are (1) convergence analysis of sparsity promoting and/or entropy reducing reconstruction schemes, (2) 3D/4D dictionary design, (3) regularization parameter choice methods for complex data noise models, and (4) convergence of intertwined iterative reconstruction schemes for problems with multiple image complexity models and/or nuisance parameters.

The above mentioned research engages four senior faculties from the Department of Mathematics at KTH and will be pursued as an integrated part of three collaborative projects described below. Each of the three postdoctoral fellows will work within a multi-professional team and take active part in the mathematical research. This calls for collaborative working skills and any experience from multi-professional teamwork is highly appreciated. Besides a strong background from computational sciences, a successful candidate must therefore have a genuine curiosity and interest towards medical imaging applications.

Project 1: Contrast enhanced CT for diagnosis & monitoring of Alzheimer's Disease (AD)
Clinical goal: Evaluate the feasibility of using CT imaging for diagnosis and monitoring of AD
Partner: The Neuro-Geriatric Memory Center at KUS.

Project 2: 4D cardiac SPECT-CT and pulmonary PET-CT imaging
Clinical goal: Increase the sensitivity/resolution of 4D cardiac SPECT-CT and pulmonary PET-CT
Partner: Department of Clinical Science, Intervention & Technology at KI

Project 3: C-arm 3D-CT for on-line treatment planning
Clinical goal: Improve the soft tissue contrast of C-arm 3D-CT imaging as to allow reliable demarcation of soft tissue tumors or to reduce the radiation exposure for precise positioning
Partner: Elekta

Description of postdoctoral fellowships

Postdoctoral fellow 1

Analytic and trained dictionaries for compression of 3D neuroanatomical and dynamic 3D cardiac and pulmonary structures

Initial task is to develop wavelet/shearlet and/or learned dictionaries relevant for representing neuroanatomical structures relevant for diagnosis of Alzheimer's Disease (project 1) and identification of brain tumors (project 3). The imaging modality here is 3D helical CT (project 1) and cone beam CT (project 3). Next, focus is on sparse representation of pulmonary and cardiac structures relevant for tumor and perfusion studies in a dynamic setting (project 2). The imaging modality is 4D emission tomography (PET and SPECT).

Postdoctoral fellow 2

Reconstruction for 4D cardiac SPECT-CT and pulmonary PET-CT imaging

This fellowship is related to project 2. Initial work is on implementing a computationally feasible SPECT-CT and  PET-CT forward model including relevant file I/O routines. Remaining part is devoted to development of variational regularization methods in collaboration with postdoctoral fellow 1.

Postdoctoral fellow 3

Reconstruction for 3D helical CT and cone-beam CT

This fellowship is related to projects 1 and 3. Initial focus is on a computationally efficient implementation of the forward model for 3D helical CT and cone-beam CT including relevant file I/O routines for correctly encoding data acquisition geometry. Next, focus is on developing TV, L1 and/or entropy type of regularization for projects 1 and 3 in collaboration with postdoctoral fellow 1.

Qualifications & eligibility

For all three fellowships, candidates should posses strong expertise in a sub-field of computational sciences (numerical mathematics, signal processing, and/or computational physics). Furthermore, candidates should have good expertise & experience from software development and some experience from cross-disciplinary research. Familiarity with the physics of radiative transport is also appreciated. The following lists specific requirements for each of the above fellowships:

Fellowship 1: Strong expertise from computational harmonic analysis, like 3D wavelet & shearlet compression and/or dictionary learning, applied to imaging problems is highly appreciated.

Fellowships 2 & 3: At least one candidate should have strong expertise from large scale non-linear optimization, preferably in the context of variational regularization. For fellowship 2, experience from TV, non-local-TV, Bregman-TV or higher order TV methods in imaging is highly appreciated. Likewise, for fellowship 3 we seek candidates with experience from L1 or entropy type of regularization in imaging.

Information about the scholarship

Form of scholarship: Time limited, up to 2 years.
Start date: According to agreement, at the latest February 1st, 2015.
Stipend compensation: According to KTH policy for postdoctoral scholarships.

Application

Application deadline: August 1, 2014
Mark your application with reference number S-2014-0390

Application must be submitted via e-mail to jobs@math.kth.se as a single pdf-file. Write the KTH reference number in the e-mail subject field, indicate the fellowship you apply for (if multiple, list them in priority order). A complete application should consist of the following parts:

  • CV
  • List of publications
  • A brief research plan
  • Copies of PhD diploma and PhD thesis
  • Contact information for three references
  • Any other relevant material that the applicant would like to be taken into account

Contacts

For more information about the project and call, consult this pd-file: and/or get in touch with the program principal investigator (Program PI) or the Head of department:

Associate Professor Ozan Öktem, Program PI
Phone: +46-733-52 21 85
E-mail: ozan@kth.se

Professor Sandra Di Rocco, Head of department
Phone: +46-8-790 7168
E-mail: dirocco@math.kth.se

Vacancies