Applications of compressive sensing in spatial frequency domain imaging

Area of Science:

• Biomedical Optics
• Medical Imaging
• Signal Processing

Background:

• Spatial Frequency Domain Imaging (SFDI) quantifies tissue optical properties using pixel-by-pixel analysis.
• Traditional SFDI can be computationally intensive and time-consuming.
• Compressive Sensing (CS) is a signal processing technique that reduces measurement requirements.

Purpose of the Study:

• To integrate CS into SFDI for improved data acquisition and analysis.
• To reduce computational time and data volume in SFDI.
• To maintain quantitative accuracy in optical property mapping.

Main Methods:

• Developed a compressive sensing SFDI (cs-SFDI) approach.
• Applied cs-SFDI to heterogeneous tissue samples (back of the hand).
• Introduced a novel CS application to the parameter recovery stage of image analysis.

Main Results:

• Achieved significant dimensionality reduction in both data acquisition and analysis.
• Demonstrated data reduction of 30% for cs-SFDI and up to 80% for parameter recovery.
• Maintained quantitative accuracy with an error of less than 10% for recovered optical properties.

Conclusions:

• CS integration in SFDI enhances efficiency without compromising accuracy.
• cs-SFDI offers advanced capabilities for multi- and hyperspectral imaging.
• This approach enables the generation of detailed optical and physiological property maps.