|
Contact
|
Michal Sofka, PhD
Image Analytics and Informatics
Siemens Corporate Research
755 College Road East
Princeton, NJ 08540
USA
Email: 
Links: Short bio,
CV in PDF format 
|
|
Research Interests
|
|
Machine learning, segmentation, object detection, registration, and matching with the application in medical imaging.
|
|
|
|
Currently, I work at
Siemens Corporate Research
in the
Image Analytics and Informatics group headed by Dr.
Dorin Comaniciu.
|
|
Previously, I did my doctoral work with professor
Chuck Stewart in the
Computer Vision Group
in the Department of Computer Science
at RPI.
|
|
Education
|
|
Rensselaer Polytechnic Institute
|
Troy, NY
|
|
Doctor of Philosophy, Computer Science
|
August 2008
|
|
Master of Science, Computer Science
|
May 2006
|
|
Union College
|
Schenectady, NY
|
|
Master of Science, Electrical Engineering
|
June 2001
|
|
Czech Technical University
|
Prague, Czech Republic
|
|
Electrical Engineering (Undergraduate Student)
|
1996 - 1998, 1999 - 2000
|
|
Union College
|
Schenectady, NY
|
|
Electrical Engineering (Exchange Student)
|
1998 - 1999
|
|
Industrial and Academic Experience
|
Siemens Corporate Research
Image Analytics and Informatics
|
Princeton, NJ
|
|
Project Manager
|
2011 - present
|
|
Managed research and development projects for various Siemens business units and external
customers. Lead team of 4-6 members. Prepared proposals for project acquisition. Monitored
targets, carried out software deliveries, and maintained customer relations. Group leader: Dorin
Comaniciu.
|
|
Research Scientist
|
2008 - 2011
|
|
Developed innovative industrial strength systems for image analysis (semantic indexing of
ultrasound anatomies and organ segmentation in CT and MR). Designed a learning-based
algorithm leveraging expert knowledge embedded in a large annotated database of medical images.
Lead development of machine learning library for object detection and segmentation. Supervised
interns. Group leader: Dorin Comaniciu.
|
Rensselaer Polytechnic Institute
Department of Computer Science
|
Troy, NY
|
|
Research Assistant
|
2003 - 2008
|
|
Implemented innovative and robust solutions to research problems
in the areas of retinal vessel extraction, uncertainty in point
and image registration, multimodal and illumination invariant
keypoint descriptors, and location registration and recognition
for mapping regions in CT scans. Published 8 papers and wrote
70K+ lines of C++ code. Supervised undergraduate research students.
|
|
Teaching Assistant
|
2001 - 2005 (4 semesters)
|
|
Led labs of 60 students in 2 groups; graded and designed homework assignments, projects, and exams; held office hours; supervised exams; and provided help outside the class.
Served as a master TA the last semester.
|
Siemens Corporate Research
Department of Intelligent Vision and Reasoning
|
Princeton, NJ
|
|
Temporary Technical Employee
|
Jan. 2004 - Aug. 2004
|
|
Designed, analyzed, validated, and maintained algorithms for automatic composing of Magnetic Resonance Images and
Angiograms. Supervised an undergraduate intern.
|
Union College
Department of Electrical Engineering
|
Schenectady, NY
|
|
Union College, Department of Electrical Engineering
|
2000 - 2001
|
|
Tutored students; provided guidance on assignments; held review sessions; and helped students to master the coursework
material. Computer assistant in the Curricular Support group.
|
|
Technical Expertise and Interests
|
|
Machine Learning: Leveraging large annotated databases; discriminative learning; effective sampling techniques; applications in image analysis.
Segmentation: Machine learning-based estimation of boundaries of structures and curve-linear segments; efficient centerline tracing; statistical shape models; interactive segmentation editing.
Detection: Multi-object detection; hierarchical detection networks.
Registration: Robust and automatic feature-based registration
in the presence of physical and illumination changes, noise,
outliers, occlusions and missing structure; uncertainty
modeling for correspondence estimation; statistical models in image registration; alignment verification.
Matching: Initialization of registration algorithms; keypoint detectors, descriptors and matching; feature extraction and scale space.
Medical imaging: Whole body image analysis; computer-assisted diagnosis and treatment planning; automatic measurements and quantification.
Data: Experience with range data (LIDAR), visible spectrum and infra red (IR) imagery, computed tomography (CT), Ultrasound (US), magnetic resonance (MR) images and angiograms (MRA).
|
|
Software Development Skills
|
-
Expert in C/C++ (multi-platform and compiler independent
development) with strong debugging and optimization skills;
proficient in Matlab.
-
Skilled in integration with software libraries: Computer Vision
Libraries, VXL (since 2001), Insight
Toolkit, ITK (since 2004), Visualization
Toolkit, VTK (since 2006).
-
Practical experience in software engineering and collaborative product development of large projects, version control systems, regression testing, continuous integration, and nightly builds.
-
Familiar with scripting languages Perl, PhP, Bash.
|
|
Mentor and Supervisor
|
|
Enrico Kuhn (Karlsruhe Institute of Technology, Germany), 2011
Vimal Singh (University of Texas, Austin, TX, USA), 2011
Jens Wetzl (University of Erlangen, Germany), 2010
Neil Birkbeck (University of Alberta, Canada), 2010
Kristof Ralovich (Technical University Munich, Germany), 2009
|
|
Professional Activities
|
|
Referee and Reviewer
|
Journals:
IEEE Transactions on Medical Imaging (TMI)
Medical Image Analysis (MedIA)
IEEE Pattern Analysis and Machine Intelligence (PAMI)
Computer Vision and Image Understanding (CVIU)
IEEE Transactions on Image Processing (TIP)
IEEE Transactions on Biomedical Engineering (TBE)
|
Conferences:
IEEE Computer Vision and Pattern Recognition (CVPR)
IEEE International Conference on Computer Vision (ICCV)
European Conference on Computer Vision (ECCV)
International Conference on Medical Image
Computing and Computer-Assisted Intervention (MICCAI)
|
|
Publications and Research
|
Please see a separate page dedicated to publications
and research projects.
|