Quantitative characterization of turbidity by radiative transfer based reflectance imaging

Area of Science:

• Biomedical Optics
• Medical Imaging
• Photonic Measurement

Background:

• Accurate optical property determination is crucial for biomedical applications.
• Noncontact, noninvasive methods are highly desirable for in vivo measurements.
• Existing techniques may lack the precision or applicability for complex tissue analysis.

Purpose of the Study:

• To develop a noncontact, inverse method for determining optical properties (absorption, scattering, anisotropy) of turbid media from a single reflectance image.
• To validate the method's ability to quantify parameters in heterogeneous targets and in vivo.
• To demonstrate the potential of quantitative reflectance imaging for melanoma diagnosis.

Main Methods:

• Development of a multispectral reflectance imaging technique.
• Utilizing a GPU-accelerated Monte Carlo simulation with a conjugate gradient descent algorithm for inverse problem solving.
• Profiling the incident beam with a diffuse reflectance standard to calibrate measurements.
• Noninvasive estimation of optical parameters in heterogeneous targets and in vivo human subjects.

Main Results:

• Successfully determined absorption coefficient (μa), scattering coefficient (μs), and anisotropy factor (g) from single reflectance images.
• Demonstrated accurate determination of embedded region thickness in heterogeneous targets.
• Provided in vivo optical parameter estimations for nevi across a spectral range (500-950nm) in four patients.

Conclusions:

• The developed quantitative reflectance imaging method offers a noncontact and noninvasive approach for determining optical properties.
• The technique shows significant potential for applications in biomedical diagnostics, specifically for melanoma detection.
• Unique solutions for optical parameters were confirmed, highlighting the robustness of the inverse method.