Three-dimensional unconstrained and constrained image-reconstruction techniques applied to fluorescence,...

Area of Science:

• Biomedical Optics
• Computational Imaging
• Medical Physics

Background:

• Near-infrared (NIR) optical imaging faces challenges in computational intensity for 3-D reconstruction and limited endogenous contrast.
• Accurate reconstruction of tissue fluorescence properties is crucial for detecting relevant biological features.

Purpose of the Study:

• To develop a computationally efficient and accurate method for 3-D inverse optical imaging.
• To recover interior fluorescence properties of exogenous contrast agents using frequency-domain photon migration measurements.
• To address the limitations of unconstrained optimization in 3-D optical imaging.

Main Methods:

• Formulated the 3-D inverse optical imaging problem as a simple-bound constrained minimization problem.
• Solved the forward optical diffusion problem using a Galerkin finite-element formulation for frustum shapes with fluorescence inclusions.
• Employed the truncated Newton method with trust regions and automatic reverse differentiation for the inverse approach.

Main Results:

• Demonstrated that constrained minimization provides a more logical approach for 3-D optical imaging compared to unconstrained methods.
• Successfully reconstructed interior fluorescence properties from boundary measurements.
• Achieved efficient computation through the use of advanced optimization techniques.

Conclusions:

• Constrained minimization is a viable and effective strategy for 3-D near-infrared optical imaging.
• The developed method enhances computational efficiency and accuracy in image reconstruction.
• This approach offers a promising solution for overcoming current limitations in biomedical optical imaging.