Optical biomarkers of neonatal hypoxic ischemic (HI) brain injury can offer the advantage of continuous, cot-side assessment of the degree of injury; research thus far has focused on examining different optical measured brain physiological signals and feature combinations to achieve this. To maximize the breadth of physiological characteristics being taken into consideration, a multimodal optical platform has been developed, allowing unique physiological insights into brain injury. In this paper we present an assessment of severity of injury using a state-of-the-art hybrid broadband Near Infrared Spectrometer (bNIRS) and Diffusion Correlation Spectrometer (DCS) instrument called FLORENCE with a machine learning pipeline. We demonstrate in the preclinical neonatal model (the newborn piglet) that our approach can identify different HI insult severity (controls, mild, severe). We show that a machine learning pipeline based on k-means clustering can be used to differentiate between the controls and the HI piglets with an accuracy of 78%, the mild severity insult piglets from the severe insult piglets with an accuracy of 90% and can also differentiate the 3 piglet groups with an accuracy of 80%. So, this analytics pipeline demonstrates how optical data from multiple instruments can be processed towards markers of brain health.
We present a new optical platform that combines broadband near-infrared spectroscopy and diffuse correlation spectroscopy for identification of brain injury severity in a preclinical model of hypoxic-ischemic encephalopathy of the neonatal brain.
We describe the development of a miniaturized broadband near-infrared spectroscopy system (bNIRS), which measures changes in cerebral tissue oxyhemoglobin ( [ HbO2 ] ) and deoxyhemoglobin ([HHb]) plus tissue metabolism via changes in the oxidation state of cytochrome-c-oxidase ([oxCCO]). The system is based on a small light source and a customized mini-spectrometer. We assessed the instrument in a preclinical study in 27 newborn piglets undergoing transient cerebral hypoxia-ischemia (HI). We aimed to quantify the recovery of the HI insult and estimate the severity of the injury. The recovery in brain oxygenation (Δ [ HbDiff ] = Δ [ HbO2 ] − Δ [ HHb ] ), blood volume (Δ [ HbT ] = Δ [ HbO2 ] + Δ [ HHb ] ), and metabolism (Δ [ oxCCO ] ) for up to 30 min after the end of HI were quantified in percentages using the recovery fraction (RF) algorithm, which quantifies the recovery of a signal with respect to baseline. The receiver operating characteristic analysis was performed on bNIRS-RF measurements compared to proton (H1) magnetic resonance spectroscopic (MRS)-derived thalamic lactate/N-acetylaspartate (Lac/NAA) measured at 24-h post HI insult; Lac/NAA peak area ratio is an accurate surrogate marker of neurodevelopmental outcome in babies with neonatal HI encephalopathy. The Δ [ oxCCO ] -RF cut-off threshold of 79% within 30 min of HI predicted injury severity based on Lac/NAA with high sensitivity (100%) and specificity (93%). A significant difference in thalamic Lac/NAA was noticed (p < 0.0001) between the two groups based on this cut-off threshold of 79% Δ [ oxCCO ] -RF. The severe injury group (n = 13) had ∼30 % smaller recovery in Δ [ HbDiff ] -RF (p = 0.0001) and no significant difference was observed in Δ [ HbT ] -RF between groups. At 48 h post HI, significantly higher P31-MRS-measured inorganic phosphate/exchangeable phosphate pool (epp) (p = 0.01) and reduced phosphocreatine/epp (p = 0.003) were observed in the severe injury group indicating persistent cerebral energy depletion. Based on these results, the bNIRS measurement of the oxCCO recovery fraction offers a noninvasive real-time biomarker of brain injury severity within 30 min following HI insult.
Broadband near-infrared spectroscopy (NIRS) can provide an endogenous indicator of tissue temperature based on the temperature dependence of the water absorption spectrum. We describe a first evaluation of the calibration and prediction of brain tissue temperature obtained during hypothermia in newborn piglets (animal dataset) and rewarming in newborn infants (human dataset) based on measured body (rectal) temperature. The calibration using partial least squares regression proved to be a reliable method to predict brain tissue temperature with respect to core body temperature in the wavelength interval of 720 to 880 nm with a strong mean predictive power of R2=0.713±0.157 (animal dataset) and R2=0.798±0.087 (human dataset). In addition, we applied regression receiver operating characteristic curves for the first time to evaluate the temperature prediction, which provided an overall mean error bias between NIRS predicted brain temperature and body temperature of 0.436±0.283°C (animal dataset) and 0.162±0.149°C (human dataset). We discuss main methodological aspects, particularly the well-known aspect of over- versus underestimation between brain and body temperature, which is relevant for potential clinical applications.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.