Info for prospective students

Hello,

Thanks for your interest in joining the lab :-)

Research in the lab

The lab focuses on developing machine learning (ML) algorithms for fast, motion-robust MRI. This includes:

• • New theory & algorithms for image reconstruction from sub-sampled data.

  • Methods for enhancing image quality, e.g. improved signal-to-noise (SNR) in low-field MRI.

  • Motion correction techniques.

  • Design of advanced sampling strategies.

  • 3D-printing of realistic phantoms.

  • AI techniques for personalized healthcare.

Desired theoretical background (coursework)

Applicants are expected to have substantial background in signal and image processing.  Specifically, the following Technion courses (or similar courses from other universities) are very useful:

• • Digital Signal Processing (DSP) course (ECE 044198) - this is essential. Alternatively, knowledge of the Fourier transform is required.

  • Image processing course (ECE 046200 or similar) 

  • Introduction to machine learning (ECE 046195 , BME 336546  or similar)

  • Deep learning course  (ECE 046211 , CS 236605 , or similar)

Desired hands-on experience:

• Substantial hands-on coding experience in Python / Matlab is required

• Experience in computer vision / deep learning (e.g. from undergrad projects or industry) is helpful

• Hands-on experience in 3D-Printing would also be great

Interested in joining?

Please email me - efrat.s at technion dot ac dot il

When applying, please make sure to attach your CV and grades transcript. Please also write some describe your previous experience in the above fields, e.g. experience from an undergraduate project.

I get many applications and I do my best to answer all of them as soon as possible. It may take a few days.. thanks for your patience.

Looking forward to hearing from you! :-)

Efrat

Relevant courses - more info

The following courses are relevant to the research in our lab:

Basic Courses in signal & image processing - highly recommended

• • Digital Signal Processing (DSP) (044198 or similar) - this is essential. Alternatively, knowledge of the Fourier transform & Nyquist sampling is essential.

  • Image processing course (ECE 046200 or similar)

Machine learning - basic courses

• • Introduction to machine learning* (ECE 046195 , BME 336546  or 096411 or similar)

  • Introduction to unsupervised learning (ECE 046202)

  • Algebraic methods for data science (095295)

  • Deep learning course  (ECE 046211 , CS 236605 , or similar)

  • Deep learning on computational accelerators (CS 236781)

Medical imaging courses

• • Introduction to medical imaging (basics of MRI, CT etc.) - BME 336502* or ECE 046831 (Guy Gilboa's course - given in spring)

  • Principles of MRI (magnetic resonance imaging) - BME 336502 (given in spring) - highly recommended

  • Introduction to medical image processing* (BME 336027)

  • Deep learning applications for MRI* (BME 336028) (Moti Freiman's course)

Optimization courses

• • Nonlinear models in operations research (096327)

  • Introduction to optimization* (Technion's 96311 or 97311 or 98311 or 236330 or 046197) 

  • Numerical algorithms* (CS 234125) 

Advanced courses in computer vision & machine learning

• Algorithms and applications in computer vision (ECE 046746)

• Diffusion Models - Miki Elad's new course* (CS 236610)

• Variational Methods in Image Processing (ECE 049064) (Guy Gilboa's course)

• Advanced Topics in Deep Learning: Transformers (ECE 046010) (Yosi Keshet's course)

*Please notice: ECE students can take only up 9 points of courses from other departments and those courses must appear in a specific list - please see here.