Time of Flight Transmission Mode Ultrasound Computed Tomography With Expected Gradient and Boundary Optimization

Area of Science:

• Medical Imaging
• Ultrasound Computed Tomography
• Biomedical Engineering

Background:

• Quantitative time of flight in transmission mode ultrasound computed tomography (TFTM USCT) offers cost-effective, non-invasive functional imaging.
• Existing TFTM USCT methods face resolution challenges due to path information concentration and transducer positioning uncertainty.
• Low-frequency TFTM USCT is particularly promising for pulmonary imaging but requires enhanced resolution and robustness.

Purpose of the Study:

• To develop and validate a novel method for enhancing resolution and robustness in low-frequency TFTM USCT for pulmonary imaging.
• To address challenges related to path information concentration and transducer positioning uncertainty in TFTM USCT.
• To improve the practical applicability and translational development of TFTM USCT for clinical lung imaging.

Main Methods:

• A new technique was developed to optimize steepest descent algorithm steps, mitigating resolution degradation from path information concentration.
• Total variation regularization was employed to stabilize the inverse problem, with a modified Barzilai-Borwein method for step size determination.
• The method was validated using simulations (k-Wave toolbox) and experimental data from a low-frequency system with a torso phantom.

Main Results:

• The proposed method significantly improved image quality in TFTM USCT reconstructions.
• Accurate retrieval of sensor locations was achieved even with imprecise initial positioning.
• The technique demonstrated robustness in realistic phantom experiments.

Conclusions:

• This study presents the first approach to address transducer location uncertainty on a transducer belt in TFTM USCT using an estimated gradient.
• The application of low-frequency USCT for lung imaging is novel, and this work tackles key practical challenges for clinical translation.
• The developed method enhances resolution and robustness, paving the way for more reliable TFTM USCT pulmonary imaging.