Referenceless Proton Resonance Frequency Thermometry Using Deep Learning with Self-Attention

TL;DR

This paper introduces a deep learning approach with self-attention for referenceless PRF MR thermometry, aiming to improve temperature monitoring accuracy during FUS treatments by addressing phase discontinuity issues.

Contribution

The study proposes a novel deep learning model with self-attention mechanisms to enhance referenceless PRF thermometry accuracy over traditional methods.

Findings

Improved temperature estimation accuracy in the presence of tissue interfaces.

Reduced errors caused by susceptibility-induced phase discontinuities.

Demonstrated robustness of the method in simulated and experimental settings.

Abstract

Background: Accurate proton resonance frequency (PRF) MR thermometry is essential for monitoring temperature rise during thermal ablation with high intensity focused ultrasound (FUS). Conventional referenceless methods such as complex field estimation (CFE) and phase finite difference (PFD) tend to exhibit errors when susceptibility-induced phase discontinuities occur at tissue interfaces.