This PDF file contains the front matter associated with SPIE Proceedings Volume 12364, including the Title Page, Copyright information, Table of Contents, and Conference Committee information.

This conference presentation was prepared for the Clinical and Translational Neurophotonics 2023 conference at SPIE BiOS, SPIE Photonics West 2023.

This conference presentation was prepared for the Clinical and Translational Neurophotonics 2023 conference at SPIE BiOS, SPIE Photonics West 2023.

This research presents our discovery of two new nonlinear optical biomarkers of Alzheimer’s disease (AD), namely 3-photon autofluorescence (3PAF), and Third Harmonic Generation (THG). A hallmark of AD is the aggregation of the Amyloid-Beta (Aβ) protein and Tau protein. Identification of these plaques and analysis of the surrounding cells and tissue is most often done using immunohistochemistry, often with inconsistent results. Using label-free nonlinear optical microscopy, new optical biomarkers were found for identifying a plaque. We present longitudinal imaging of AD progression in mice ranging from 8 to 52 weeks in age, in the hippocampal and cortical regions.

Neurosurgery is the first line treatment for most malignancies of the brain however intraoperative healthy and diseased tissue differentiation often remains a challenge. We have demonstrated earlier that wide-field Muller Polarimetry Imaging (MPI) is a promising approach for brain tissue differentiation and fiber tracking. To examine the technique’s versatility in a similar to in vivo setting, we used our system to create maps of polarimetric properties for tissue differentiation in cadaveric animal brains under neurosurgery-like conditions. We present the effects of ultrasonic cavitation on optical response and examined the challenges of a complex topography and blood presence in a surgical resection cavity.

The purpose is to determine whether diffuse reflectance spectroscopy (DRS) can provide optical guidance during deep brain stimulation (DBS) surgery. Experiments on monkey ex vivo brains have been performed to ensure DRS methods could differentiate white and gray matter. In this study, we use principal component analysis (PCA) to determine the composition of tissue in front of the stimulation electrode. Furthermore, our work tackles the mechanical consequences of implementing an optical probe in a DBS electrode. This multidisciplinary project shows that DRS can be used as a non-invasive, cost-effective and real-time tissue characterization.

Diffuse correlation spectroscopy (DCS) offers non-invasive measurements of tissue perfusion and is increasingly broadly applied in human subject research, in particular in the neuromonitoring arena. However, signal to noise (SNR) limitations have prompted great interest in alternative instrumentation approaches to address this issue, such as the speckle contrast optical spectroscopy (SCOS) technique which uses spatial multi-speckle contrast to estimate blood flow. Here we present a simulation study of the brain perfusion sensitivity achievable by each method on adults, to guide the use of SCOS vs DCS approaches in future studies. We find that SCOS brain sensitivity is comparable to DCS.

Motor execution, observation, and imagery are basic skills that are often used in motor learning and rehabilitation. Although unimodal neuroimaging studies had suggested potential neural mechanisms underlying action execution, observation, and imagination, the results are inconsistent. Using a multimodal approach (i.e., simultaneous recording of functional near-infrared spectroscopy (fNIRS) and Electroencephalogram (EEG)), we confirmed consistent activation over the left inferior parietal lobe, superior marginal gyrus, and post-central gyrus during all three conditions. Our multimodal findings suggested concise regions associated with the action observation system and provides new insights into interpreting unimodal findings.

We piloted long-term diffuse correlation spectroscopy (DCS) monitoring of cerebral blood flow in a patient with an aneurysmal subarachnoid hemorrhage. Measurements were conducted for 18 days. We also recorded blood pressure, ECG, and other clinical monitors as available. The blood flow index from 5, 25, and 30 mm separation channels showed a variety of responses depending on the patient condition and treatment. As an example, repeated doses of nimodipine were given for treatment purposes, resulting in level or increased cerebral blood flow despite a decrease in mean arterial blood pressure. There was correlation between the short-distance channel and heart rate.

There are gaps in our understanding of the neural signatures underlying naturalistic deficits associated with impaired motor imitation in autism spectrum disorder (ASD) due to logistical limitations in neuroimaging modalities like fMRI. Therefore, we utilized high-density diffuse optical tomography (HD-DOT) to image twenty-three young adults as they observed and imitated sequences of upper extremity movements. Alterations in multiple cortical areas were observed when comparing neural responses to motor observation and motor imitation in this sample. This establishes the utility of HD-DOT for neuroimaging during naturalistic overt motion

Machine learning algorithms require a large and diverse data set for robust training. However, gathering a sufficient number is a difficult task due to time and budget constraints. Generated data sets can augment training data and provide diverse example for training. We propose a method to generate realistic diffuse optical tomography (DOT) data sets based on known physiological components of the DOT signal. We generate three dimensional models of each signal component and seed the hemodynamic response to activate targeted cortices. Our method reduces the need for a large recruitment process and increases the accuracy of machine learning algorithms.

High-density diffuse optical tomography (HD-DOT) has been shown to be a promising alternative to fMRI for mapping cortical hemodynamics in young healthy adults. HD-DOT imaging can be more precise when coupled with subject specific head models rather than generic atlas-based head models. While MRI-derived head models are commonly used, in some patient groups including subjects with metal and/or electrical implants, only CT images can be obtained. In this study, we developed a CT-based head modeling pipeline and demonstrated the feasibility of improved mapping of brain responses to tasks compared to a generic atlas-based head-model.

Surgical skill assessment requires the development of objective metrics for improved training regimen and certification process. In this project, we sought to leverage fNIRS data within the framework of the Fundamental of Laparoscopy program to 1) classify the subject's expertise level (n=16), 2) monitor surgical skill acquisition (n=36), 3) benchmark against and/or predict the FLS score which is used for surgical certification (n=13). The data set is made open access to enable neuroscience discovery, provide test beds for improved data analysis and provide data to support the development of AI-based solutions in fNIRS.

Transcatheter aortic valve replacement (TAVR) surgery has a risk of cognitive impairment and neurological injury. Currently, there are few options for non-invasively monitoring brain activity and perfusion, with electroencephalography, transcranial Doppler, and near-infrared spectroscopy (NIRS) all having significant drawbacks. By combining NIRS with diffuse correlation spectroscopy (DCS) we can obtain a more complete picture of cerebral hemodynamics during TAVR procedures and examine the link to neurological outcomes. We show examples of post-valve replacement hemodynamic changes that correspond with worse/better patient outcomes

This study addresses the pressing need for non-invasive, continuous monitoring of cerebral physiologic derangements following traumatic brain injury (TBI). We combine frequency-domain and broadband diffuse optical spectroscopy with diffuse correlation spectroscopy to monitor cerebral oxygen metabolism, cerebral blood volume, and cerebral water content in an established adult swine-model of focal TBI. Cerebral physiology is monitored before and after TBI (up to 14 days post injury). Overall, our results suggest that non-invasive optics can monitor cerebral physiology impairments such as reduced oxygen metabolism, hemorrhage, and edema formation post-TBI.

Cardiac surgery with cardiopulmonary bypass (CPB) is associated with postoperative neurological complications. Targeted mean arterial blood pressure (MAP) during cardiac surgery is used as one method of maintaining adequate cerebral blood flow (CBF) and perfusion pressure. However, an MAP target of 60 mmHg after transitioning on CPB, which is used in many centers, does not account for the reported broad range of lower autoregulatory limits (50-90 mmHg) [1]. In an effort to maintain cerebral perfusion, near-infrared spectroscopy (NIRS) is used to monitor tissue oxygen saturation (StO₂); however, StO₂ is not a direct marker of CBF or tissue oxygen demand. As an alternative, possible effects on cerebral energy metabolism could be monitored by using hyperspectral NIRS (hsNIRS) to measure changes in the redox state of cytochrome c oxidase (ΔoxCCO), which are linked to ATP production. In this study, an in-house built hsNIRS/diffuse correlation spectroscopy (DCS) was used to monitor ΔoxCCO, CBF and StO₂ in patients during cardiac surgery with CPB. Fourteen patients were retrospectively grouped according to the level of their MAP when transitioning onto CPB: high (70-90 mmHg), target (57-69 mmHg), and low MAP (40-56 mmHg). The aim was to evaluate the potential effects of MAP on ΔoxCCO during the transition onto CPB. Results demonstrated that the smallest changes in oxCCO (-0.08 ± 0.24 μM) were observed in the high MAP group and significantly larger changes (-0.73 ± 0.25 μM) in the low MAP group. The results highlight the potential of ΔoxCCO monitoring for real-time assessment of MAP management during CPB with the ultimate aim of mitigating adverse cerebral events.

Traumatic brain injury (TBI) is the most severe type of injury in terms of death and lifelong disability. The mainstay of severe TBI management is intracranial pressure (ICP) measurement. Existing gold standard techniques to measure ICP involve placing a sensor into the brain tissue through a small hole drilled in the skull. This non-therapeutic procedure risks infection and bleeding into the brain and can only be performed by a neurosurgeon. Therefore, there is a vital demand to develop non-invasive technologies that allow measuring the ICP.

This research has developed a non-invasive, continuous, and optical monitoring system to assess ICP in TBI patients. An in-vitro evaluation of the technology was performed in a head phantom that mimics cerebral anatomy, physiology, and optical tissue properties. The system works by shining light into the brain through the skull. The recorded optical signals (photoplethysmograph) from the phantom’s brain are related to the induced changes in ICP. In this setup, an invasive ICP probe was also used as the reference measurement. The phantom’s pressure was changed between 5 mmHg and 30 mmHg in steps of 5 mmHg. The experiment considered 16 replicates. Advanced algorithms and Machine Learning (ML) models utilising optical signal feature extraction techniques were implemented in translating the optical signals into absolute measurements of ICP. The resultant support vector machine model presented a sensitivity and specificity of 87% ± 10 and 92% ± 13, respectively. This proof of concept study suggests the system’s viability for performing non-invasive estimations of ICP.

Intracranial pressure (ICP) measurements help monitor patient status following cerebral injury, and currently require implantation of an invasive pressure probe. The potential complications associated with this implantation have restricted the application of ICP measurements in less severe conditions. We propose a non-invasive alternative that derives features from the cardiac waveforms present in near-infrared spectroscopy (NIRS) measurements and inputs these features into a decision tree regressor to estimate ICP. We evaluated this method in nine subjects already fitted with invasive ICP sensors. The non-invasive nature of NIRS instrumentation eases the clinical adoption of this ICP estimation approach.

We use diffuse correlation spectroscopy (DCS) to measure changes in microvascular blood flow in subarachnoid hemorrhage patients treated with intrathecal nicardipine. Our results suggest that IT nicardipine achieves the desired effect of microvascular vasodilation in the majority of patients. The CBF response plateaus by day 3, indicative of steady state of drug concentration in the brain. Interestingly, those patients whose microvascular cerebral blood flow did not respond were the ones who went on to develop worse outcome in the form of a secondary stroke.

In children with sickle cell disease, there is a clinical need for non-invasive quantification of the degree of hemometabolic stress in these patients to mitigate risk of stroke. Frequency-domain near-infrared spectroscopy (FDNIRS) and diffuse correlation spectroscopy (DCS) measures of regional oxygen extraction fraction, cerebral blood flow, and cerebral metabolic rate of oxygen have potential to provide markers of cerebral metabolic stress. In this study, we characterize the intra-subject and inter-operator repeatability of these measures, and we correlate DCS measures of cerebral blood flow index against both arterial spin-labeled MRI and transcranial Doppler ultrasound in a cohort of pediatric SCD patients.

Cerebral autoregulation (CA), the brain’s ability to regulate perfusion independently of blood pressure, can be assessed by evaluating the degree of correlation between cerebral blood flow (CBF) and mean arterial pressure. Non-invasive, optical measurements of brain hemodynamics using DCS/NIRS can be used to assess CA and show agreement with invasive metrics (laser doppler perfusion and intracranial pressure) in a pediatric swine model of cardiac arrest. Wavelet based coherence methods of assessing autoregulation are a useful alternative to correlation based methods.

Cerebral autoregulation (CA) is a mechanism to maintain cerebral blood flow (CBF) in response to changes in cerebral perfusion pressure (CPP), through active vasoconstriction and vasodilation of arterioles in the brain. Dynamic CA is believed to act as a high-pass filter such that only low frequency changes in pressure are counteracted by an active vasculature response. With high frequency oscillations in pressure, such as those that occur at the heart rate (HR), the effects of dynamic CA are absent and changes in CPP are passively transmitted to CBF based on the cerebrovascular resistance (CVR) and compliance (CVC). These changes in CVR/CVC occur with steady-state changes in CA which can be described by Lassen’s curve. However, it is unclear what drives phase differences between pressure and flow at the respiration rate of around 0.2 Hz (12 breaths per minute). Quantifying phase differences at the physiologic respiration rate could be useful to gain a better understanding of the effects of CA and as a potential clinical monitoring tool. In this work, we looked at phase differences between arterial blood pressure (ABP) and intracranial pressure (ICP) measured with invasive pressure sensors, which serve as surrogates for CPP and CBF, to investigate how Arg(ABP)-Arg(ICP) change at the respiration rate as a function of the CPP. We quantify how Arg(ABP)-Arg(ICP) changes with respect to CPP after low-frequency oscillations, respiratory induced oscillations, and with oscillations driven by the heart rate. In each frequency regime, the trends in phase differences between Arg(ABP)-Arg(ICP) are unique with respect to CPP. At the respiration rate, the trends in Arg(ABP)-Arg(ICP) did not completely follow those predicted by a dynamic CA response or by CVC/CVR, thus we believe that there is a combination of effects influencing the phase difference between Arg(ABP)-Arg(ICP) at the respiration frequency. We also explore whether this response could be monitored completely non-invasively using near infrared spectroscopy (NIRS). We use Arg(ΔHbT)-Arg(ΔHbO) as surrogates for CPP and CBF and see a similar response of phase differences with respect to CPP at the respiration rate.

Bedside tools are needed to alert clinicians to the onset of ischemic stroke. Resting-state hemodynamics with optical intrinsic signal imaging (OIS) were assessed in mice before and after middle cerebral artery (MCA) stroke. OIS analysis included resting-state functional connectivity (FC), low frequency power (LFP, local brain activity), and temporal-shift delay (impaired perfusion). Immediately after stroke, there is a decrease in homotopic connectivity and an absence of LFP in stroke-affected hemisphere; perfusion deficits were more localized. Over 24 hours, LFP and delayed perfusion localized in the core MCA territory (matching TTC staining). Such biomarker development could translate to optical bedside technologies.