Generalized conditional gradient method with adaptive regularization parameters for fluorescence molecular...

Area of Science:

• Biomedical Optics
• Medical Imaging
• Computational Biology

Background:

• Fluorescence molecular tomography (FMT) enables in vivo 3D visualization of fluorescent probes.
• FMT reconstruction faces challenges due to light scattering and ill-posed inverse problems.
• Existing methods struggle with balancing sparsity, shape preservation, and robustness.

Purpose of the Study:

• To enhance the performance of FMT reconstruction.
• To develop a robust and accurate method for imaging fluorescently labeled probes in vivo.
• To address limitations of traditional regularization techniques in FMT.

Main Methods:

• Proposed a generalized conditional gradient method with adaptive regularization parameters (GCGM-ARP).
• Introduced elastic-net (EN) regularization to balance sparsity and shape preservation.
• Utilized the L-curve for adaptive regularization parameter adjustment.
• Employed generalized conditional gradient method (GCGM) for efficient sub-problem solving.

Main Results:

• GCGM-ARP demonstrated superior performance in numerical simulations and in vivo experiments.
• Achieved minimum location error (LE) and relative intensity error (RIE).
• Obtained the maximum Dice coefficient (Dice) across various conditions, including noise.
• Showcased enhanced source localization, dual-source resolution, morphology recovery, and robustness.

Conclusions:

• GCGM-ARP is an effective and robust strategy for FMT reconstruction.
• The method significantly improves upon existing mathematical reconstruction techniques.
• Offers superior accuracy and reliability for biomedical applications of FMT.