Marcus Björk – PhD Student

Research

I'm a PhD student at the division of Systems and Control under Professor Peter Stoica. I started January 2010 after completing my engineering physics MSc. thesis in the same division. My research is mainly in the fields of Signal Processing and System Identification and is funded by an ERC advanced grant focusing on biomedical applications. I've been involved in a project treating identification of systems with quantized output. Particularily on modeling the human response to countiouos Levodopa treatment of Parkinson's disease. The goal is to find more efficient treatment strategies.
My current research is in Magnetic Resonance Imaging (MRI). By modeling the MR signal and estimating the parameters from data, one can extact intresting information such as the magnetic field inhomogeneitiy which can be used to remove dark bands in the images and improve the SNR. By modeling we can also separate fat and water in the images or find the temperature. The problem is to design efficient parameter estimation algorithms that solve these, genereally nonlinear, problems.

Publications

1. SNR Requirements for T1 and T2 Estimation using bSSFP. R. Reeve Ingle, Joëlle K. Barral, Marcus Björk, Erik Gudmundson, Peter Stoica, and Dwight G. Nishimura. In Proceedings of the ISMRM 20th annual meeting, 2012.
2. Optimality of Equally-Spaced Phase Increments for Banding Removal in bSSFP. Marcus Björk, R. Reeve Ingle, Joëlle K. Barral, Erik Gudmundson, Dwight G. Nishimura, and Peter Stoica. In Proceedings of the ISMRM 20th annual meeting, 2012.
3. Signal Modeling and the Cramér-Rao Bound for Absolute Magnetic Resonance Thermometry in Fat Tissue. Marcus Björk, Johan Berglund, Joel Kullberg, and Peter Stoica. In Proc. 45th Asilomar Conference on Signals, Systems, and Computers, pp 80-84, 2011. (DOI).
4. Signal Processing Algorithms for Removing Banding Artifacts in MRI. Marcus Björk, Erik Gudmundson, Joëlle Barral, and Peter Stoica. In Proceedings of the 19th European Signal Processing Conference (EUSIPCO-2011), European Signals Processing Conference, pp 1000-1004, 2011. (fulltext).
5. Dynamic models with quantized output for modeling patient response to pharmacotherapy. Marcus Björk, Alexander Medvedev, and Peter Stoica. In Proc. International Conference on Control Applications: CCA 2010, pp 1029-1034, IEEE, Piscataway, NJ, 2010. (DOI).
6. Estimation of dynamic models with output quantization applied to drug response modeling. Marcus Björk and Alexander Medvedev. Reglermöte 2010, Lund, 2010.

Links

Transformmetoder 2012

Teaching

HT12

• Period 2, responsible for W:s projects in Transformmetoder given by Bo Styf at the Math department.
• Period 2, lab assistant in Reglerteknik I (F3) given by Magnus Lundberg Nordenvaad at the IT department.

2011

• Period 2, responsible for W:s projects in Transformmetoder given by Bo Styf at the Math department.

• Period 1, lab assistant in Empirisk modellering (ES5, STS5, ES5) given by Bengt Carlsson at the IT department.

• Period 4, exercise sessions in Modellering av Dynamiska System (IT, STS) given by Bengt Carlsson at the IT department.

• Period 3, lab assistant in Reglerteknik I (W3, ES3) given by Hans Norlander at the IT department.

2010

• Period 2, responsible for IT, STS, X and W:s projects in the two courses in Transformmetoder given by Lennart Salling (for IT, W and X) and Anders Södergren (for STS), at the Math department.

• Period 1, lab assistant in Reglerteknik II (F4, STS4, W4) given by Torsten Söderström at the IT department.