Digital instrument simulator to optimize the development of hyperspectral systems: application for intraoperative...

Area of Science:

• Neuroscience
• Biomedical Engineering
• Optical Imaging

Background:

• Intraoperative optical imaging assesses brain function by monitoring hemodynamics and metabolism.
• Current methods struggle with quantification errors due to cortical optical property variations and empirical spectral configuration choices.
• Accurate intraoperative brain mapping is crucial for neurosurgical procedures.

Purpose of the Study:

• To develop a digital instrument simulator for optimizing hyperspectral imaging systems for intraoperative brain mapping.
• To identify optimal wavelengths for monitoring cerebral hemodynamics (hemoglobin) and metabolism (cytochromes).

Main Methods:

• Utilized white Monte Carlo simulations of exposed cortex from real images.
• Employed a genetic algorithm-based optimization procedure to find optimal wavelength combinations in the visible and near-infrared spectrum.
• Modeled intensity maps and path length images to account for optical property inhomogeneities.

Main Results:

• The simulator identified an 18-wavelength combination reducing quantification errors for oxygenated hemoglobin, deoxygenated hemoglobin, and oxidized cytochrome c oxidase by 47%, 57%, and 57%, respectively.
• This optimized set significantly outperformed the gold standard of 121 wavelengths.
• Improvements in resolving cytochrome changes were limited, but oxidized cytochrome c oxidase resolution was enhanced.

Conclusions:

• A digital instrument simulator and optimization framework were developed for hyperspectral systems in intraoperative brain mapping.
• This approach can guide the design of more accurate and efficient hyperspectral imaging devices for neurosurgery.