Toward a generic real-time compression correction framework for tracked ultrasound.

Area of Science:

• Medical Imaging
• Biomechanical Modeling
• Surgical Navigation

Background:

• Tissue compression during ultrasound imaging introduces significant errors in subsurface target localization and geometry.
• Accurate visualization is critical for effective soft tissue interventions and surgical guidance.

Purpose of the Study:

• To present a novel, generic tissue compression correction method for ultrasound imaging.
• To enable real-time, near-video framerate compression compensation independent of preoperative data.

Main Methods:

• A generic block tissue model calibrated to the ultrasound probe tip was developed.
• Intraoperative surface digitization measured compression, providing boundary conditions for a biomechanical model.
• The model's displacement field was inverted to nonrigidly transform ultrasound images, correcting for compression.

Main Results:

• In phantom studies, the method reduced mock tumor margin error (MHD) from 5.0 ± 1.6 to 2.1 ± 0.7 mm.
• Centroid alignment error in phantoms decreased from 7.6 ± 2.6 to 2.6 ± 1.1 mm.
• Clinical application reduced in vivo tumor margin MHD error from 5.4 ± 0.1 to 2.9 ± 0.1 mm and centroid error from 7.2 ± 0.2 to 3.8 ± 0.4 mm.

Conclusions:

• The developed compression correction method effectively improves ultrasound image alignment with tomographic data.
• This generic approach offers improved efficiency compared to previous patient-specific correction methods.