A Novel Adaptive Parameter Search Elastic Net Method for Fluorescent Molecular Tomography.

Area of Science:

• Biomedical Imaging
• Medical Physics
• Computational Biology

Background:

• Fluorescence molecular tomography (FMT) enables in vivo quantitative reconstruction of fluorescent probes.
• Traditional Lp norm regularization in FMT suffers from over-sparseness, over-smoothness, and poor robustness.

Purpose of the Study:

• To develop an improved regularization method for FMT reconstruction.
• To enhance the accuracy, resolution, and robustness of FMT imaging.

Main Methods:

• Proposed an adaptive parameter search elastic net (APSEN) method combining L1 and L2 norms.
• Introduced L0 and L2 norms for adaptive selection of elastic net weight parameters.
• Validated APSEN using digital mouse models and in vivo liver tumor experiments.

Main Results:

• APSEN demonstrated superior location accuracy, spatial resolution, and fluorescence yield recovery.
• The method exhibited enhanced morphological characteristics and robustness against noise and parameter variations.
• In vivo experiments confirmed the practical applicability of APSEN for FMT.

Conclusions:

• The APSEN method offers significant improvements over traditional techniques for FMT reconstruction.
• APSEN enhances quantitative accuracy and robustness in complex biological imaging scenarios.
• This advancement holds promise for more precise molecular imaging in preclinical and clinical settings.