welcome back to another episode of beyond the patterns so today i had
the great pleasure to welcome mertz a bunker to our video here so mert
received a phd in electrical engineering from princeton
university where his dissertation dealt with biomedical image
registration martin moved to the massachusetts institute of
technology for a postdoc at computer science and artificial intelligence
lab where he worked on a range of biomedical image analysis problems
including the segmentation of brain mri scans after his post doc at mit
mert was a faculty member of the a martino center for biomedical
imaging in the massachusetts general hospital and harvard medical school
where he built a research program on algorithmic tools for integrating
large scale genetics and medical imaging datasets today de merde is
associate professor in the school of electrical and computer engineering
at cornell and cornell tech in new york city his research group
devolves machine learning based and computational tools for biomedical
image applications and he is all also the recipient of an nsf career
award that he was awarded in two thousand and eighteen in an nih
early career development award in two thousand and eleven so murder is
going to present us today really cool stuff about compressed sensing
and how to integrate deep learning of that you will see he has really a
lot of smarter ideas so i'm really glad that he's here to eva face but
you'll also see is that i actually have been traveling why we did the
recording so i have been in a hotel room in montreal canada so it's
really great that we can travel again and you'll see that now and
back and we can finally upload this video and you will be able to enjoy
a merge presentation so mert without further ado the stage is yours and
thank you andrea thanks for the info it's great to be here i just
thinking when's the last time i gave a ballpark like this and it's been
awhile about almost a year now so i might be a little rusty but here it
goes out today i'm going to talk about contrast imaging or compressed
sensing for imaging and before i die there and i want to and there's a
couple pr slides and i just recently moved to new york city and where are
we cornell university has a couple campuses are brand new campus our
newest campus in your city is called cornell tech it says that unique
graduate education campus with a sort of a societal impact and
outreach mission and my primary lab space and office spaces are at it's
campus it's on roosevelt island for those of you who are familiar with
new york city it's on roosevelt island right next to manhattan and
we have a wonderful views of the manhattan skyline and and and yeah
so this is sort of what it looks like it's construction just recently
the birthplace of the construction completed like three years ago i
think right now and we have another phase that coming up in five years
and cornell also has a medical campus in new york it's on the upper
east side of manhattan on sixty ninth then first and york and and
it's called weill cornell medicine so we have the medical school and we
have the ability at the hospital which is near miss parian and i also
have an affiliation with the medical school and i also have a lap space
at the medical campus as well so that was sort of a little bit of pr
i said just a couple of disclaimers i consult for a company called
clearly none of the work i'm going to talk about i think today involves
Presenters
Zugänglich über
Offener Zugang
Dauer
01:19:23 Min
Aufnahmedatum
2021-11-07
Hochgeladen am
2021-11-07 13:16:04
Sprache
en-US
We have the great honor to welcome Mert Sabuncu to our lab for an invited presentation!
Abstract: Imaging techniques such as MRI can be accelerated by sampling below the Shannon-Nyquist rate via compressed sensing. In this talk, I will consider the use of deep learning methods for this problem. First, I will present our approach for Learning-based Optimization of the Under-sampling PattErn, or LOUPE. For a given sparsity constraint, LOUPE optimizes the under-sampling pattern and reconstruction model simultaneously, using a computationally efficient end-to-end deep learning strategy. Our experiments with MRI and microscopy demonstrate that LOUPE-derived patterns yield significantly more accurate reconstructions compared to standard under-sampling schemes. I will then switch gears and focus on the reconstruction problem only and presents some deep-learning based innovations that we have recently proposed, including the use of hyper-networks that give end-users to ability to choose from multiple reconstructions that are consistent with data.
Short Bio: Mert R. Sabuncu received a PhD degree in Electrical Engineering from Princeton University, where his dissertation dealt with biomedical image registration. Mert then moved to the Massachusetts Institute of Technology for a post-doc at the Computer Science and Artificial Intelligence Lab, where he worked on a range of biomedical image analysis problems, including the segmentation of brain MRI scans. After his post-doc at MIT, Mert was a faculty member at the A.A Martinos Center for Biomedical Imaging (Massachusetts General Hospital and Harvard Medical School), where he built a research program on algorithmic tools for integrating large-scale genetics and medical imaging datasets. Today, Mert is Associate Professor in the School of Electrical and Computer Engineering at Cornell University and Cornell Tech, in New York City. His research group develops machine learning based computational tools for biomedical imaging applications. He is a recipient of an NSF CAREER Award (2018) and an NIH Early Career Development Award (2011).
Register for more upcoming talks here!
References
Bahadir, Cagla Deniz, Adrian V. Dalca, and Mert R. Sabuncu. "Learning-based optimization of the under-sampling pattern in MRI." International Conference on Information Processing in Medical Imaging. Springer, Cham, 2019.
Bahadir, Cagla D., et al. "Deep-learning-based optimization of the under-sampling pattern in MRI." IEEE Transactions on Computational Imaging 6 (2020): 1139-1152.
Zhang, Jinwei, et al. "Extending LOUPE for K-space Under-sampling Pattern Optimization in Multi-coil MRI." International Workshop on Machine Learning for Medical Image Reconstruction. Springer, Cham, 2020.
This video is released under CC BY 4.0. Please feel free to share and reuse.
For reminders to watch the new video follow on Twitter or LinkedIn. Also, join our network for information about talks, videos, and job offers in our Facebook and LinkedIn Groups.
Music Reference:
Damiano Baldoni - Thinking of You (Intro)
Damiano Baldoni - Poenia (Outro)