Meeting Banner
Abstract #3381

SeqGPT: Training a Large Language Model to Generate MRI Pulse Sequences

Snawar Hussain1, Jörn Huber1, Matthias Günther1, and Daniel Christopher Hoinkiss1
1Fraunhofer MEVIS, Bremen, Germany

Synopsis

Keywords: Language Models, Acquisition Methods

Motivation: LLMs possess remarkable ability to process and generate sequential data which can be used to generate MRI pulse sequences.

Goal(s): Train a custom implementation of Generative Pre-trained Transformer (GPT) to generate MRI pulse sequences.

Approach: Pseudo random version of sequences like FLASH were generated to create training set which were then tokenized through a custom tokenizer with reduced vocabulary size. These tokenized sequences were used to train a custom implementation of GPT.

Results: GPT was able to learn the sequence structure and generate out-of-data pseudo random sequences that can be simulated to get a reconstruction.

Impact: This shows the capability of LLMs to generate MRI sequences. Which then can be fine-tuned with a differentiable simulator to adjust the sequence towards desired imaging objectives.

How to access this content:

For one year after publication, abstracts and videos are only open to registrants of this annual meeting. Registrants should use their existing login information. Non-registrant access can be purchased via the ISMRM E-Library.

After one year, current ISMRM & ISMRT members get free access to both the abstracts and videos. Non-members and non-registrants must purchase access via the ISMRM E-Library.

After two years, the meeting proceedings (abstracts) are opened to the public and require no login information. Videos remain behind password for access by members, registrants and E-Library customers.

Click here for more information on becoming a member.

Keywords