ESMRMB Wiki
ESMRMB Home Wiki Main Page Job Board Conference Lab

Self-learning MR: Difference between revisions

← Older edit
VisualWikitext
Accepted
No edit summary
 
(2 intermediate revisions by the same user not shown)
Line 24: Line 24:
   | Speaker2= Or Perlman
   | Speaker2= Or Perlman
   | Speaker2_Confirmed= yes
   | Speaker2_Confirmed= yes
}}
|Speaker3=Moritz Zaiss|Speaker4=Shaihan Malik|Speaker3_Confirmed=Yes|Speaker4_Confirmed=Yes}}


<!-- Add the full description of the session here below -->
<!-- Add the full description of the session here below -->
Line 32: Line 32:
In recent years, learning-based reconstruction and post-processing have transformed clinical MRI—making scans faster, sharper, and more informative. However, as data processing continues to improve, the focus is shifting back to data acquisition. With smaller artifacts becoming more apparent, we need to fix them again in acquisition. With reconstruction methods extracting all available information, the frontier of generation of new contrast features gets a revival.  
In recent years, learning-based reconstruction and post-processing have transformed clinical MRI—making scans faster, sharper, and more informative. However, as data processing continues to improve, the focus is shifting back to data acquisition. With smaller artifacts becoming more apparent, we need to fix them again in acquisition. With reconstruction methods extracting all available information, the frontier of generation of new contrast features gets a revival.  


This session addresses all kind of automized sequence discovery or optimization approaches -  from historic EPG-based flip angle train optimization, to pulse design, and MRF schedules - trying to put it under the name "Self-learning MR". Yet, self-learning MR has many names: Optimal Control, End-to-End Optimization, Known-Operator Learning, Physics-Informed AI, Strategy Discovery, Optimal Design of Experiments, Dynamic Inverse Problems  
This session addresses all kind of automized sequence discovery or optimization approaches -  from historic EPG-based flip angle train optimization, to MRF schedules - trying to put it under the name "Self-learning MR". Yet, self-learning MR has many names: Optimal Control, End-to-End Optimization, Known-Operator Learning, Physics-Informed AI, Strategy Discovery, Optimal Design of Experiments, Dynamic Inverse Problems  


Potential topics could be:
While learning-based methods are well-known on the reconstruction side, we explicitly focus here on the MR preparation/acquisition side in the broadest sense. Of course, some tailored acquisitions can go hand in hand with tailored reconstruction.
Self-learning MRF / CEST MRF / MRS
Self-learning trajectories
Self-learning contrast, sharpness
Self-learning over a real scanner
Self-learning imperfection mitigation
Self-learning pulses, pTX/SMS pulses
Self-learning flip angle trains, phases


While learning-based methods are well-known on the reconstruction side, we explicitly focus here on the MR preparation/acquisition side in the broadest sense. Of course, some tailored acquisitions can go hand in hand with tailored reconstruction.
All materials are shared via github: https://github.com/mritogether/ESMRMB2025_Self_learning_MR


Speakers:  
The program will be the following:
Tony Stöcker - classic EPG-based optimizations of TSE, SFFP. pTX pulse design.
Or Perlman - Overview on MRF, autoseq, and autoCEST.
Moritz Zaiss – End-to-end sequence design


We already have an interest group mailing list with 19 groups/people, who I could invite to submit an abstract if this session gets chosen.
=== Program for first hour: 4 talks ===
 
The program will be the following:


Option 1: (1:30h)
* Welcome and Definition of the session (Focus on Acquisition, beyond k-space patterns and recon) (Shaihan and Moritz) 2 mins
* '''Tony Stöcker, DZNE, Bonn: “EPG-based signal modeling and sequence optimization”''' (15 min)
* '''Shaihan Malik, KCL, London: “Parallel Transmit, from fields to excitations to sequences”''' (15 min)
* '''Moritz Zaiss, FAU, Erlangen: “End-to-end learning with MR-zero”''' (15 min)
* '''Or Perlman, Tel Aviv University, “Towards Automatically Optimized Multi-Metabolite CEST Fingerprinting – When a Computational Graph Meets Proton Exchange"''' (15 min)


*Presentation by Tony Stöcker on history of EPG-based optimization + advances +  pTx ''(20 min)''
=== Program for the second half an hour: demos of open source “self-learning” tools ===
* Presentation by Moritz Zaiss on end to end learing ''(10 min)''
* Presentation by Or Perlmann on AutoCEST''(10 min)''
* Selected ESMRMB Abstracts that fit to the topic: ''(50 min)''


* '''Svenja Niesen  "Optimize it! EPGs in Python"''' (5 min)  Paper: Availabe soon.  Code:<nowiki>https://github.com/mrphysics-bonn/epg-tutorial</nowiki>
* '''Jonathan Endres – Pulseq-zero''' (5 min):  Paper: <nowiki>https://github.com/pulseq-frame/pulseq-zero/blob/main/abstract/abstract.md</nowiki> Code: <nowiki>https://github.com/pulseq-frame/pulseq-zero</nowiki>  DSC + DESC
* '''Felix Glang – Trajectory Optimization''' (5 min):


Option 2:  (1h)
Paper: <nowiki>https://archive.ismrm.org/2021/4200.html</nowiki> Code:


* Presentation by Tony Stöcker on history of EPG-baased optimization + advances ''(15 min)''
* '''Dario Bosch – fast pTX''' (5 min) Paper: <nowiki>https://doi.org/10.1007/s10334-023-01134-7</nowiki> Code: <nowiki>https://github.com/dabosch/FastPtx</nowiki>
* Presentation by N. Boulant or Tony Stöcker on Ptx pulses ''(10 min)''
* '''Nikita Vladimirov: – CEST MRF for everyone''' (5 min) Paper:  <nowiki>https://doi.org/10.1038/s41596-025-01152-w</nowiki>  Code:  <nowiki>https://github.com/momentum-laboratory/molecular-mrf</nowiki>
* Presentation by Moritz Zaiss on end to end learing ''(15 min)''
* Virtual:
* Presentation by Or Perlmann on AutoCEST''(15 min)''
** '''Jannik Stebani/Martin Blaimer - Optimization of quantification using Cramer-Rao Lower bound''' (5 min) Paper: Code:<nowiki>https://github.com/HarmonizedMRI/OpenMRF</nowiki>
* Discussion ''(5 min)''

Latest revision as of 14:50, 29 September 2025



Self-learning MR

Sequence Discovery & Optimization - where we come from, where we go.

Session duration: 60m - 90m
Intended audience: Physicists, Radiologists, AI-folks
Moderators: Moritz Zaiss, Shaihan Malik (Confirmed: Zaiss, Malik)
First Speaker: Tony Stöcker (Confirmed: yes)
Second Speaker: Or Perlman (Confirmed: yes)
Third Speaker: Moritz Zaiss (Confirmed: Yes)
Fourth Speaker: Shaihan Malik (Confirmed: Yes)



Session description

In recent years, learning-based reconstruction and post-processing have transformed clinical MRI—making scans faster, sharper, and more informative. However, as data processing continues to improve, the focus is shifting back to data acquisition. With smaller artifacts becoming more apparent, we need to fix them again in acquisition. With reconstruction methods extracting all available information, the frontier of generation of new contrast features gets a revival.

This session addresses all kind of automized sequence discovery or optimization approaches - from historic EPG-based flip angle train optimization, to MRF schedules - trying to put it under the name "Self-learning MR". Yet, self-learning MR has many names: Optimal Control, End-to-End Optimization, Known-Operator Learning, Physics-Informed AI, Strategy Discovery, Optimal Design of Experiments, Dynamic Inverse Problems

While learning-based methods are well-known on the reconstruction side, we explicitly focus here on the MR preparation/acquisition side in the broadest sense. Of course, some tailored acquisitions can go hand in hand with tailored reconstruction.

All materials are shared via github: https://github.com/mritogether/ESMRMB2025_Self_learning_MR

The program will be the following:

Program for first hour: 4 talks

  • Welcome and Definition of the session (Focus on Acquisition, beyond k-space patterns and recon) (Shaihan and Moritz) 2 mins
  • Tony Stöcker, DZNE, Bonn: “EPG-based signal modeling and sequence optimization” (15 min)
  • Shaihan Malik, KCL, London: “Parallel Transmit, from fields to excitations to sequences” (15 min)
  • Moritz Zaiss, FAU, Erlangen: “End-to-end learning with MR-zero” (15 min)
  • Or Perlman, Tel Aviv University, “Towards Automatically Optimized Multi-Metabolite CEST Fingerprinting – When a Computational Graph Meets Proton Exchange" (15 min)

Program for the second half an hour: demos of open source “self-learning” tools

  • Svenja Niesen "Optimize it! EPGs in Python" (5 min) Paper: Availabe soon. Code:https://github.com/mrphysics-bonn/epg-tutorial
  • Jonathan Endres – Pulseq-zero (5 min): Paper: https://github.com/pulseq-frame/pulseq-zero/blob/main/abstract/abstract.md Code: https://github.com/pulseq-frame/pulseq-zero DSC + DESC
  • Felix Glang – Trajectory Optimization (5 min):

Paper: https://archive.ismrm.org/2021/4200.html Code:

  • Dario Bosch – fast pTX (5 min) Paper: https://doi.org/10.1007/s10334-023-01134-7 Code: https://github.com/dabosch/FastPtx
  • Nikita Vladimirov: – CEST MRF for everyone (5 min) Paper: https://doi.org/10.1038/s41596-025-01152-w Code: https://github.com/momentum-laboratory/molecular-mrf
  • Virtual:
    • Jannik Stebani/Martin Blaimer - Optimization of quantification using Cramer-Rao Lower bound (5 min) Paper: Code:https://github.com/HarmonizedMRI/OpenMRF
Retrieved from "https://wiki.esmrmb.org/index.php?title=Self-learning_MR&oldid=1226"