We use cookies to help provide you with the best possible online experience. Learn more


Post-doctoral position with KTH

KTH Royal Institute of Technology in Stockholm has grown to become one of Europe’s leading technical and engineering universities, as well as a key centre of intellectual talent and innovation. We are Sweden’s largest technical research and learning institution and home to students, researchers and faculty from around the world. Our research and education covers a wide area including natural sciences and all branches of engineering, as well as in architecture, industrial management, urban planning, history and philosophy.

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

Job description 

The Department of Mathematics at KTH is offering two-year post-doctoral positions within the Brummer & Partners MathDataLab. The Brummer & Partners MathDataLab is a research lab in mathematics and applied mathematics, hosted at the Department of Mathematics, that aims at creating a hub for mathematical research in the analysis of complex data. The Brummer & Partners MathDataLab supports top talent among junior researchers by offering competitive postdoc positions, and aims to stimulate collaboration between existing faculty and surrounding institutions. A further aim is to inspire and facilitate the education of a new generation of mathematicians focused on foundations and implementations of novel methods for data analysis.

The position is time limited, full time, 1+1 year postdoc positions starting July 1, 2018 or according to agreement.


A PhD degree, awarded (or planned to be awarded before the commencement of the position) in mathematics, applied mathematics, computer science or related areas is a requirement. We seek a candidate with a strong background in parts of mathematics relevant to the research activity of one or more of the proposed projects. The successful applicant should be strongly motivated, have the capability to work independently as well as in collaboration with members of the research group, and have good communication skills. No skills in Swedish are required.

The project: A statistical machine learning approach to tumour target volume estimation

Henrik Hult, Professor in Mathematical Statistics, Department of Mathematics, KTH
Atsuto Maki, Associate Professor, RPL, School of Computer Science and Communication, KTH
Industry partner: RaySearch Laboratories AB

Radiotherapy plays a central role in the treatment of patients with cancer. Roughly 30-40% of the population will develop cancer, and at least half of them require radiotherapy at some time during their illness. Most of the time radiotherapy is given with curative intent, often in combination with surgery and chemotherapy.

During the last two decades, treatment planning for radiotherapy has evolved from forward planning, where dose is calculated from manually created beam setups, to inverse planning, where the beam setup and delivery parameters are optimised based on dose-based objective functions reflecting the goals of the treatment. However, creating a treatment plan that has an acceptable trade-off between the conflicting clinical goals, such as target coverage, dose conformity around the target and organ at risk sparing using inverse planning, is still a highly manual and time-consuming process. The main reasons for this is that the objective function does not directly correspond to all the clinical goals of the treatment and that the internal patient geometry differs between patients. Since the geometry differs between patients, the achievable trade-offs between dose to different structures will differ, which means that the same objective function cannot be applied to patients even though the tumour type is identical.

The tumour target volume is a central quantity in the planning and treatment process. There are three main target volumes in radiotherapy planning. The first is the location and extent of the gross tumour which often can be imaged, this is the gross tumour volume (GTV). The second volume, the clinical target volume (CTV), contains the GTV plus a margin for tumour cell spread that cannot be imaged. The third volume is the planning target volume (PTV) which is the CTV plus a margin to account for uncertainties in the treatment planning and delivery. Although the target volume is conceptually easy, in practice, the edges of the target volume are not always clear and different imaging modalities (e.g. computer tomography, magnetic resonance imaging, etc.) may contribute different aspects of the target volume estimation. In particular, the CTV is very hard to identify from image data which may lead to large variations in the volume definition between doctors and clinics. Today much data, in the form of image sets, is stored for patients treated with radiotherapy.

The goal of this project is to create a framework for automatic estimation of the tumour target volume based on image data. For instance, given a GTV the framework should estimate the CTV. Statistical techniques and machine learning techniques such as convolutional neural networks could be combined with prior beliefs and clinical information to approach this important problem. An extension of the project would be to apply the framework to organ volumes as well.

This is a joint project with RaySearch Laboratories AB. For this project we are primarily looking for a candidate with a background in mathematical statistics and/or computer science.

About RaySearch Laboratories: 

RaySearch is advancing cancer treatment through pioneering software. We believe software has unlimited potential, and that it is now the driving force for innovation in oncology. RaySearch works in close cooperation with leading cancer centers to bring scientific advancements faster to the clinical world. Today, our solutions support thousands of clinics worldwide in the fight against cancer. By making oncology software faster, easier and more flexible, we enable better care for cancer patients worldwide.


Log into KTH's recruitment system in order to apply to this position ny blicking on the button below. You are the main responsible to ensure that your application is complete according to the ad.

Your complete application must be received at KTH no later than 15.Dec.2017 11:59 PM CET.

Letters of recommendation can also be sent directly by the letter writers to jobs@math.kth.se.