Extended Kalman filtering for continuous volumetric MR-temperature imaging

Area of Science:

• Medical Physics
• Biomedical Engineering
• Radiology

Background:

• Real-time magnetic resonance (MR) thermometry is crucial for guiding high-intensity focused ultrasound (HIFU) therapies.
• Current MR-thermometry methods face limitations in temporal/spatial resolution, signal-to-noise ratio (SNR), and physiological noise, impacting precision.
• MR-guided HIFU requires volumetric temperature maps for comprehensive monitoring of heating trajectories and thermal effects.

Purpose of the Study:

• To develop and evaluate an advanced MR-thermometry technique for improved volumetric temperature monitoring during HIFU interventions.
• To enhance the accuracy, precision, and spatial/temporal resolution of real-time temperature measurements.
• To address limitations posed by under-sampled data and physiological noise in MR-guided thermal therapies.

Main Methods:

• A multi-slice MR sequence was used for continuous scanning of the targeted area during HIFU heating.
• An extended Kalman filter (EKF) integrated measured MR data with a priori 3D data from a physics-based model forecast.
• The method was validated through ex vivo phantom experiments and in vivo porcine kidney studies.

Main Results:

• The proposed reconstruction method significantly improved temperature measurement resolution and precision while ensuring output accuracy.
• In vivo experiments demonstrated successful reconstruction of 3D volumetric temperature data from spatio-temporally under-sampled datasets.
• The method achieved a high temporal resolution (0.127 s) for a large field of view (142.5×285×54 mm³) with reduced noise and minimal latency.

Conclusions:

• The developed EKF-based MR-thermometry technique offers a robust solution for real-time volumetric temperature monitoring in MR-guided HIFU.
• This approach enhances the reliability and effectiveness of HIFU therapies by providing accurate and precise thermal feedback.
• The method holds significant potential for advancing image-guided thermal ablation and other MR-guided interventions.