Nonlinear approach to difference imaging in diffuse optical tomography.

Area of Science:

• Biomedical Optics
• Medical Imaging
• Optical Physics

Background:

• Difference imaging recovers changes in optical properties using pre- and post-change measurements.
• Conventional methods use linearized difference reconstruction, offering good tolerance to modeling errors but limited quantitative accuracy and spatial resolution.
• Linearization of nonlinear observation models can lead to qualitative difference images with weak spatial resolution.

Purpose of the Study:

• To develop and evaluate a nonlinear approach for difference imaging that overcomes the limitations of traditional linear methods.
• To enable simultaneous reconstruction of optical parameters before and after a change using two datasets.
• To assess the feasibility and robustness of the nonlinear approach against various modeling errors.

Main Methods:

• A nonlinear difference imaging method was developed for simultaneous reconstruction of optical parameters from two measurement datasets.
• The approach was validated using numerical simulations.
• Experimental data from a phantom study were used to test the method's performance and error tolerance.

Main Results:

• The nonlinear approach demonstrated improved quantitative accuracy and spatial resolution compared to linearized methods.
• The method showed feasibility in recovering changes in optical properties.
• The study investigated the approach's tolerance to common modeling errors, including domain truncation, optode coupling errors, and domain shape errors.

Conclusions:

• The developed nonlinear difference imaging method offers a promising alternative to conventional linear techniques.
• Simultaneous reconstruction of pre- and post-change optical parameters enhances image quality and reliability.
• The nonlinear approach exhibits robustness against several types of modeling errors, suggesting its potential for practical applications in biomedical optics.