Waveform inversion with source encoding for breast sound speed reconstruction in ultrasound computed tomography

Area of Science:

• Medical Imaging
• Biomedical Engineering
• Acoustics

Background:

• Ultrasound computed tomography (USCT) shows potential for breast cancer detection and management.
• Waveform inversion (WI) methods offer superior spatial resolution compared to ray-based methods in USCT.
• Conventional WI methods are computationally intensive, limiting their application in USCT.

Purpose of the Study:

• To develop an accelerated USCT reconstruction method for breast imaging.
• To address the computational burden of conventional waveform inversion techniques.
• To maintain high spatial resolution in USCT images while reducing computational cost.

Main Methods:

• Developed the waveform inversion with source encoding (WISE) method for USCT.
• Employed source encoding concepts to accelerate the reconstruction process.
• Utilized a stochastic optimization problem with a stochastic gradient descent algorithm to estimate sound speed distribution.
• Validated the WISE method using computer simulations and experimental phantom studies.

Main Results:

• The WISE method significantly reduces the computational burden of USCT reconstruction.
• WISE maintains the high spatial resolution properties characteristic of waveform inversion methods.
• Both simulation and phantom studies demonstrated the efficacy of the WISE method.

Conclusions:

• The WISE method offers an efficient and effective approach for USCT breast imaging.
• This accelerated method makes advanced waveform inversion techniques more accessible for clinical applications.
• WISE has the potential to improve breast cancer detection and management through enhanced imaging capabilities.