Referenceless Proton Resonance Frequency Thermometry Using Deep Learning with Self-Attention
This addresses the need for more reliable temperature monitoring in medical procedures like high-intensity focused ultrasound ablation, though it appears incremental as it builds on existing referenceless methods with a new deep learning approach.
The paper tackled the problem of errors in proton resonance frequency MR thermometry caused by susceptibility-induced phase discontinuities at tissue interfaces during thermal ablation monitoring, by developing a referenceless method using deep learning with self-attention, which improved accuracy compared to conventional techniques.
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.