Steady-State Free Precession Relaxation Mapping with Deep Learning in High-Resolution NMR

Acronym

P_SSFP-DL

Consortium Coordinator

Crizostomo Kock, Flavio Vinicius

Start Date

September 1, 2025

End Date

August 28, 2026

Status

https://purl.org/pe-repo/concytec/estadoProyecto#activo

Tipo de proyecto

https://purl.org/pe-repo/ocde/tipoProyecto#investigacionBasica

Description

Steady-State Free Precession (SSFP) in Nuclear Magnetic Resonance (NMR) spectroscopy is a powerful technique for acquiring high-resolution spectra. However, accurately determining longitudinal (T1) and transverse (T2) spin relaxation times from SSFP data presents a significant challenge due to the complex interplay between signal intensity and numerous experimental parameters. To address this challenge, this research proposes a novel deep learning (DL)-driven framework for directly mapping simultaneous T1 and T2 values from SSFP datasets. This approach overcomes the limitations of traditional methods, such as inversion recovery and Carr-Purcell-Meiboom-Gill (CPMG) pulse sequence, by employing deep learning routines to model spin dynamics and accurately map relaxation times. These architectures will be trained to learn the complex relationships between experimental parameters (flip angle (θ), RF pulse duration, repetition time (TR), echo time (TE)), SSFP signal intensities, and the underlying relaxation times (T1 and T2). Furthermore, this DL-based approach offers several key advantages, including enhanced accuracy and efficiency, improved robustness against experimental noise and artifacts, and increased flexibility for adapting to diverse SSFP pulse sequences and experimental conditions. Consequently, this research has the potential to significantly advance NMR spectroscopy by enabling more accurate and efficient relaxation time measurements using SSFP, thereby accelerating research across diverse fields, including biomolecular NMR, materials science, and metabolomics.

Keywords

Resonancia magnética nuclear

;

Aprendizaje profundo

;

Mapeo de relajación

;

Alta resolución

Área de conocimiento

Natural sciences

Campo OCDE

https://purl.org/pe-repo/ocde/ford#1.04.07