Blind source unmixing in multi-spectral optoacoustic tomography

Area of Science:

• Biomedical Imaging
• Optical Imaging
• Medical Physics

Background:

• Multispectral optoacoustic tomography (MSOT) combines imaging depths of optical methods with ultrasound resolution.
• MSOT can identify tissue biomarkers by decomposing spectral contributions of photo-absorbing molecules.
• Accurate decomposition is crucial for quantitative analysis and biomarker identification in MSOT.

Purpose of the Study:

• To investigate the performance of blind source unmixing and spectral fitting for decomposing fluorescent dye contributions in MSOT.
• To evaluate these methods for spectral unmixing in fluorescence imaging.
• To demonstrate the capacity of temporal unmixing for real-time MSOT data in vivo.

Main Methods:

• Acquisition of MSOT measurements in mice.
• Application of blind source unmixing algorithms.
• Application of spectral fitting approaches.
• Analysis of real-time, in-vivo MSOT data for temporal unmixing.

Main Results:

• Blind source unmixing demonstrated promising accuracy for MSOT decomposition.
• The method is suitable for spectral unmixing in fluorescence imaging.
• Temporal unmixing effectively enhanced visualization of absorber agent flow in the mouse vascular system.

Conclusions:

• Blind source unmixing is a robust technique for spectral decomposition in MSOT.
• This approach holds potential for improved biomarker identification and imaging in biological tissues.
• The study highlights the utility of blind unmixing for advanced MSOT applications, including real-time in-vivo imaging.