Evaluating online filtering algorithms to enhance dynamic multispectral optoacoustic tomography.

Area of Science:

• Biomedical optics
• Medical imaging
• Otoacoustic imaging

Background:

• Multispectral optoacoustic tomography (MSOT) offers high spatiotemporal resolution imaging.
• Acquiring multispectral images requires sequential wavelength scanning, risking aliasing and noise.
• Dynamic spectral changes can be obscured by noise in time-series MSOT data.

Purpose of the Study:

• To develop and validate novel filtering techniques for MSOT time-series data.
• To mitigate inter-frame noise and improve image quality in dynamic MSOT acquisitions.
• To enhance downstream analyses like spectral unmixing and quantitative physiological measurements.

Main Methods:

• Application of kinematic filters to multispectral MSOT time-series data.
• Validation using a multispectral Shepp-Logan phantom.
• In vivo testing on mice with renal cell carcinoma tumors undergoing an oxygen challenge.

Main Results:

• Kinematic filters effectively suppressed inter-frame noise in dynamic MSOT image time courses.
• Demonstrated improved signal-to-noise ratio in both phantom and in vivo experiments.
• Enabled higher-fidelity spectral unmixing and more sensitive detection of spectral changes.

Conclusions:

• Kinematic filtering is a valuable tool for enhancing MSOT data quality.
• The improved signal fidelity facilitates more accurate quantitative analysis of biological processes.
• This approach holds promise for advancing diagnostic capabilities using MSOT.