An In Silico Exploration of Aberration-Corrected Transspinal Focused Ultrasound Using Zero Echo Time MRI.

Area of Science:

• Medical Imaging
• Ultrasound Technology
• Biophysics

Background:

• Focused ultrasound (FUS) offers non-invasive therapeutic potential.
• Accurate focusing through bone, like the spine, is challenging due to acoustic property variations.
• Magnetic resonance imaging (MRI) guidance is desirable for FUS but requires accurate acoustic property mapping.

Purpose of the Study:

• To numerically assess the feasibility of MRI-corrected transspinal focused ultrasound (FUS).
• To compare the accuracy of different correction methods using CT and MRI-derived data.
• To determine the impact of acoustic property representation on focusing accuracy.

Main Methods:

• Ex vivo thoracic vertebrae were imaged using CT and zero echo time (ZTE) MRI.
• ZTE images were converted to pseudo-CT (pCT) for acoustic property estimation.
• Transspinal FUS was simulated with five correction strategies, including no correction, pCT-corrected (homogeneous/heterogeneous), and CT-corrected (homogeneous/heterogeneous).
• Simulated pressure fields were evaluated against a CT-corrected heterogeneous gold standard.

Main Results:

• MRI-corrected focusing (pCT) improved spatial accuracy and target contour similarity compared to no correction.
• pCT-corrected focusing with heterogeneous properties reduced spatial shift (1.5 ± 2.0 mm) and improved Dice similarity (0.75 ± 0.10).
• Homogeneous CT-corrected focusing closely approximated the gold standard, with minimal spatial shift (0.8 ± 1.7 mm) and high Dice similarity (0.91 ± 0.06).

Conclusions:

• MRI-corrected transspinal FUS is feasible and outperforms uncorrected methods.
• Accurate morphological representation of the spine is critical for effective MR-corrected FUS.
• While promising, current MRI-based corrections do not fully match the accuracy of CT-based corrections for transspinal FUS.