Michael J. Allison

Ph.D. Candidate
Department of Electrical Engineering and Computer Science (EE:S)
University of Michigan
1301 Beal Avenue
Ann Arbor, MI 48109

I am a Ph.D. candidate in the Electrical Engineering: Systems program within the Department of Electrical Engineering and Computer Science (EECS) at the University of Michigan. I work on magnetic resonance (MR) image reconstruction under the guidance of Prof. Jeffrey A. Fessler. My research projects have included compressive sensing in MRI, receive coil sensitivity estimation, and field map estimation. My work has been supported in part by the Natural Sciences and Engineering Research Council of Canada (NSERC) and the NIH.

Research

I have explored the following MRI topics during my Ph.D. career:

Sensitivity Map Estimation
- • Proposed a regularized algorithm for estimating the sensitivity maps of MR surface coils both within and around the object support (ISMRM 2011 and Manuscript).
- • Developed and analyzed algorithms to accelerate the computation of these estimates (ISMRM 2011 and ISBI 2012).

Field Map Estimation
- • Created several algorithms to accelerate the computation of regularized field map estimates which requires the minimization of a non-convex cost function (ISMRM 2012 Magna Cum Laude Award).

Compressive Sensing in MRI
- • Explored the effects of Poisson disk sampling and minimum sample spacing on compressive sensing MR reconstruction.
- • Found interesting thresholding behavior when evaluating the image reconstruction quality versus minimum sample spacing.

I have also explored computer vision and machine learning topics through course projects:

Three-Dimensional Generic Object Representation
- • Recreated the work of S. Savarese et al. on 3D generic object categorization, localization, and pose estimation.
- • Used to satisfy part of the course requirements for EECS 598 -- Special Topics in Computer Vision (Prof. S. Savarese).

Confidence Region Estimation for Medical Image Registration
- • Developed a machine learning algorithm that estimated a confidence region for each pixel in a set of registered medical images.
- • The work was greatly extended by a fellow group member into a conference paper.
- • Used to satisfy part of the course requirements for EECS 545 -- Machine Learning (Prof. C. Scott).

Education

University of Michigan, Ann Arbor, USA

Ph.D. Candidate, Electrical Engineering: Systems, Sept. 2010 - Present, ( Expected Winter 2014 ).
M.S.E., Electrical Engineering: Systems, Sept. 2008 - Apr. 2010.

University of Waterloo, Canada

B.A.Sc., Systems Design Engineering, Honours, Co-operative, Sept. 2003 - June 2008.
Alumni Gold Medal in Engineering

Selected Publications

[1]	Michael J. Allison, Sathish Ramani, and Jeffrey A. Fessler. Accelerated regularized estimation of MR coil sensitivities using augmented Lagrangian methods. IEEE Transactions on Medical Imaging, vol. 32, no. 3, pages 556-564, March 2013.
[2]	Michael J. Allison and Jeffrey A. Fessler. Accelerated computation of regularized field map estimates. In Proceedings of International Society of Magnetic Resonance in Medicine , page 413, Melbourne, May 2012.
[3]	Michael J. Allison, Sathish Ramani, and Jeffrey A. Fessler. Regularized MR coil sensitivity estimation using augmented Lagrangian methods. In Proceedings of IEEE International Symposium on Biomedical Imaging , Barcelona, May 2012.
[4]	Michael J. Allison and Jeffrey A. Fessler. An augmented Lagrangian method for MR coil sensitivity estimation. In Proceedings of International Society of Magnetic Resonance in Medicine , page 2881, Montreal, May 2011.
[5]	Khaled Shaban, Abdunnaser Younes, Robert Lam, Michael Allison, and Shajeehan Kathirgamanathan. Credit default swap pricing using artificial neural networks. In Proceedings of IEEE 2010 International Joint Conference on Neural Networks (IJCNN 2010) , pages 1467-1474, Barcelona, July 2010.

* Please refer to my CV for a complete list.

Major Awards and Honors

2012	—	ISMRM Magna Cum Laude Merit Award, Melbourne.
2011 - 2013	—	Natural Sciences and Engineering Research Council (NSERC) PGS-D Fellowship.
2008 - 2010	—	Natural Sciences and Engineering Research Council (NSERC) PGS-M Fellowship.
2008 - 2009	—	Electrical Engineering and Computer Science Departmental Fellowship.
2008	—	Ontario Graduate Scholarship (declined).
2008	—	University of Waterloo Alumni Gold Medal in Engineering.
2008	—	Sandford Fleming Foundation Academic Achievement Award.
2007 - 2008	—	Ontario Professional Engineers Foundation Undergraduate Scholarship.
2005 - 2006	—	Microsoft Corporation Merit Scholarship.
2005	—	Harold E. Johns Summer Studentship in Medical Physics.
2003 - 2004	—	J.N.E. Consulting Ltd. Entrance Scholarship.
2003	—	Governor General's Bronze Medal.

* Please refer to my CV for a complete list.

Work Experience

I have held one internship during my graduate career as a summer associate with McKinsey & Company based out of the Toronto office.

I have also completed six internships during my undergraduate degree as part of the co-operative education program. These included research positions at the London Regional Health Sciences Centre in London, ON and the Sunnybrook Health Research Centre in Toronto, ON. I was a software developer/trading assistant on the trade floor of TD Securities in Toronto, ON. I also worked as a software analyst at JSI Telecom in Kanata, ON and a technical assistant at the University of Waterloo.

* Please refer to my CV for additional details.

Teaching

Graduate Student Instructor, EECS 451: Digital Signal Processing, University of Michigan, Fall 2010.

Design Projects

The Systems Design Engineering program places a significant emphasis on hands-on product development experience. Some of the more interesting projects I participated in were:

Augmented Reality for Art Galleries

Group members: Kevin Po and Sheel Shah

This was my senior thesis developed over the 2007 - 2008 academic year. Our goal was to design an augmented reality system that would help guide a museum patron through art exhibits while displaying relevant information through a heads-up display.

The final device used a webcam and a pair of augmented reality glasses to survey the user's surroundings and overlay information onto the artwork in real-time. To assist in detecting paintings in low light, we used infrared LEDs as markers. We then used a modified version of the SURF algorithm to identify each painting and its orientation. A major challenge of the project was achieving real time speed on our hardware.

I would do many things differently after taking several computer vision courses. However, I found this project fun and the resulting prototype was very impressive considering the time frame. Here is a short summary with details.

SureJect — An Easy to Carry Epinephrine Injection System

Group members: Sujeev Ruban, Eunice Ho, and Kaveh Mirsaeidi

A major limitation of existing epinephrine injection devices is that they require the user to carry a large syringe at all times. Our goal was to design a smaller injection device that could be inserted into the spine of a wallet. Thus, you would not have to think about carrying a separate unit.

This project required a significant amount of mechanical modeling and prototype testing to ensure that the injection forces and volumes were within tolerances. We built the injection system out of aluminum and our working prototype fit into a standard men's wallet. I learned a great deal about precision machining from this project.

Sturdy Lift

Group members: Kevin Po, Sheel Shah, Sujeev Ruban, Emmanuel Elmajian, and Sunit Mohindroo

The goal was to design a device to assist the elderly around the home. After interviewing residents of a local long-term care facility, we decided to create a device that could help them reach objects in the cupboard. The resulting design was to modify a rollator (walker) with an electric platform that could raise a person two feet off the ground.

We built a prototype using a donated walker and solicited feedback from our original participants. The reviews were positive, but I was hesitant to let them try it. In hindsight, I think this project had flaws from the beginning. However, it was a good learning experience from both the human factors and engineering perspective and it was featured in the local newspaper.

If you are still interested, here is the poster with the original CADD images.

Curriculum Vitae

CV — Updated Dec 16, 2013

Resume — a two page summary