Improving the quantitative accuracy of cerebral oxygen saturation in monitoring the injured brain using atlas based Near Infrared Spectroscopy models

Abstract

The application of Near Infrared Spectroscopy (NIRS) for the monitoring of the cerebral oxygen saturation within the brain is well established, albeit using temporal data that can only measure relative changes of oxygenation state of the brain from a baseline. The focus of this investigation is to demonstrate that hybridisation of existing near infrared probe designs and reconstruction techniques can pave the way to produce a system and methods that can be used to monitor the absolute oxygen saturation in the injured brain. Using registered Atlas models in simulation, a novel method is outlined by which the quantitative accuracy and practicality of NIRS for specific use in monitoring the injured brain, can be improved, with cerebral saturation being recovered to within 10.1 ± 1.8% of the expected values.

Reliability of fNIRS for noninvasive monitoring of brain function and emotion in sheep

“…Further, differently from what it is normally done in fNIRS studies on humans where probe placement can be guided by proper neuro-anatomical images and functional atlas 33 – 35 , the placement of the probe on the animal head was not aided by magnetic resonance imaging (MRI) data and rarely it was not consistently supported by neuro-anatomical or functional atlas of the animal brain 25 . Consequently, fNIRS signals were attributed to the frontal or pre-frontal region, also hypothesizing a vascular stealing mechanism in the brain cortex 23 , 26 – 28 , but without determining the sensitivity of the technique to that specific cortical region.…”

Functional near infrared spectroscopy using spatially resolved data to account for tissue scattering: A numerical study and arm‐cuff experiment

“…The use of multiple overlapping measurements has demonstrated the ability of NIR spectroscopy (NIRS) to image the brain region using tomographic reconstruction techniques [8]. These methods can use the anatomical information of different regions of the head derived from other imaging modality such as an MRI, and predict the photon propagation using numerical modeling of the diffusion approximation to model the light intensities for every source-detector combination and iteratively fit the optical properties of the medium such that the modeled data matches with the measured data [9]. As these methods take into consideration the head geometry, tomographic imaging methods tend to give more accurate measure of cerebral optical properties as compared to homogeneous infinite medium approximations.…”

Special Issue: Selected papers from the Topical Problems of Biophotonics conference

“…In vivo monitoring of oxygen saturation in the brain by NIR spectroscopy provides unique opportunities for non‐invasive brain disorder diagnostics and monitoring; thus, Clancy et al. present MRI‐based numerical simulation of NIR spectroscopy towards improving the accuracy of quantitative mapping of oxygen saturation in brain.…”