Influence of absorption and scattering on the quantification of fluorescence diffuse optical tomography using...

Area of Science:

• Biomedical optics
• Medical imaging
• Fluorescence imaging

Background:

• Near-infrared fluorescence imaging often normalizes light intensity, assuming uniform tissue optical properties.
• Tissue heterogeneity in absorption and scattering can significantly impact quantification accuracy.

Purpose of the Study:

• Investigate the impact of absorption and scattering variations on quantification in near-infrared fluorescence imaging.
• Evaluate reconstruction improvements using heterogeneous models compared to homogeneous ones.

Main Methods:

• Developed computer-simulated phantoms with varying scattering and absorption properties.
• Included homogeneous, heterogeneous slab, and atlas-based mouse liver/lung phantoms.
• Compared quantification errors between homogeneous and heterogeneous reconstruction models.

Main Results:

• Assuming homogeneous media led to significant quantification errors (+41% to +94% for scattering, -7% to +44% for absorption/complex phantoms).
• Errors were highly sensitive to scattering coefficient variations when assuming homogeneity.
• Heterogeneous models reduced overall quantification errors to -7% to +7%.

Conclusions:

• Homogeneous models in near-infrared fluorescence imaging introduce substantial quantification errors due to optical property variations.
• Employing heterogeneous models significantly enhances quantification accuracy, crucial for reliable imaging.