Time-of-flight resolved light field fluctuations reveal deep human tissue physiology

The study of light paths in random media is crucial in various fields, including biophotonics. Near-infrared spectroscopy (NIRS) aims to assess deep biological tissues in vivo by measuring highly scattered near-infrared light. NIRS methods can be categorized as incoherent (CW-NIRS, frequency-domain, and time-domain NIRS) or coherent (DWS/DCS, laser Doppler, and laser speckle). Coherent NIRS, particularly DWS/DCS, probes RBC dynamics related to blood flow. While the theory connecting scatterer dynamics to light fluctuations is established, classical DWS/DCS methods remain largely empirical, lacking a definitive physical model for RBC motion. This disconnect stems from the inability of classical DWS/DCS to resolve light dynamics in time-of-flight (TOF) and the fact that TOF-integrated intensity autocorrelations are similar for Brownian motion and random flows at intermediate source-detector separations. The researchers aim to advance in vivo diffuse light scattering experiments by introducing interferometric near-infrared spectroscopy (iNIRS), a technique capable of directly measuring intrinsic field autocorrelations with high TOF resolution. The improved methodology will help to resolve existing ambiguity in the interpretation of blood flow from classical NIRS techniques and advance the measurement of deep tissue physiology.

The literature review highlights the limitations of existing near-infrared spectroscopy (NIRS) techniques, specifically diffuse correlation spectroscopy (DCS), in accurately measuring deep tissue blood flow. Classical DCS methods struggle to resolve the time-of-flight (TOF) of scattered light and often yield autocorrelations that resemble Brownian motion even in the presence of random flow, leading to difficulties in distinguishing between competing physical models. Previous studies have employed numerical simulations to investigate the effects of different flow models (Brownian motion vs. advection), but the resolution of in vivo measurements has been insufficient to definitively resolve these models. The authors cite several studies that employ other approaches, such as coherence-gated methods, to try to isolate singly scattered light, but these techniques are limited to superficial tissue measurement. The existing limitations of classical DCS emphasize the need for a more sophisticated approach like iNIRS, which offers improved temporal and spatial resolution.

The researchers employed interferometric near-infrared spectroscopy (iNIRS) to measure time-of-flight (TOF)-resolved light field fluctuations in human tissues. iNIRS measures the interference spectrum of light passing through tissue and a reference path, yielding a mutual coherence function and TOF-resolved optical field autocorrelations. A crucial aspect of the methodology involved correcting for motion artifacts. They used short source-detector separations to obtain a static reference signal for estimating and correcting phase drift caused by bulk sample motion. Digital spectral shaping was used to maintain dynamic range. The correction methodology was validated against optical coherence tomography (OCT) measurements, demonstrating excellent agreement (R² = 0.95). The TOF-resolved autocorrelations were then analyzed using various fitting models: a 3-parameter exponential model (static and dynamic terms), a 5-parameter bi-exponential model, and a 2-parameter model for late TOFs. These models allowed for the quantification of decay rates and amplitudes of different components in the autocorrelations. The study involved measurements in an Intralipid phantom, human forearm, nude mouse head, and human forehead. Physiological manipulations, such as Intralipid injection, hypercapnic challenge, and cognitive tasks, were performed to assess the sensitivity of iNIRS to blood flow changes. Monte Carlo simulations were also conducted to investigate the influence of different scattering properties and flow models on the shape of autocorrelations. Finally, the researchers compared iNIRS results with classical DWS/DCS measurements to evaluate the effectiveness of iNIRS in deconstructing and interpreting classical DCS data.

The iNIRS technique, after motion correction, successfully revealed two significant findings: First, the TOF-resolved autocorrelations exhibited bi-exponential decay across various tissues and experimental conditions, except in Intralipid (a homogeneous scattering medium). The presence of a slow dynamic component in addition to the fast one, suggests sampling media with different dynamics. This bi-exponential behavior is particularly prominent at early to intermediate TOFs and is attributed to 'dynamical snake paths', light paths characterized by predominantly small-angle scattering events in intravascular space. This contrasts with the exponential decay typically observed in homogeneous media, confirming the complex nature of light transport in biological tissues. The researchers’ simulations support this observation. The amplitude of the slower component relates to the scattering anisotropy and shape of the scattering phase function. The simulations also highlight how the slow component’s amplitude depends on sample dynamics and intravascular forward scattering. Second, the study demonstrated the capability of iNIRS to deconstruct and interpret classical DWS/DCS results. By integrating the fast component of the iNIRS autocorrelations, consistent results were obtained with DWS/DCS theory. Moreover, the iNIRS method provides a way to determine blood flow indices without explicitly measuring absorption, which is required for accurate interpretation of classical DWS/DCS measurements. The study also observed a transition to higher decay rates at later TOFs in the human forehead, consistent with differences in blood flow between superficial and deep tissues. The Monte Carlo simulations corroborated this biphasic behavior in a two-layer model of human forehead, highlighting the capability of iNIRS to distinguish between superficial and deep dynamics. The results from physiological manipulations consistently showed changes in the decay rates and amplitudes of the autocorrelations, demonstrating the sensitivity of iNIRS to blood flow dynamics.

The findings of this study directly address the limitations of classical diffuse optical flowmetry (DOF) techniques and provide a more refined understanding of light-tissue interactions. The introduction of TOF resolution in iNIRS allows for a direct confirmation of the Brownian nature of RBC dynamics, resolving the ambiguity present in previous studies that relied on TOF-integrated autocorrelations. The identification of ‘dynamical snake paths’ offers a novel explanation for the complex decay patterns observed in biological tissues. Furthermore, the ability of iNIRS to deconstruct and accurately interpret classical DWS/DCS measurements is significant, improving the accuracy and reliability of blood flow assessments. The distinct capability of iNIRS to distinguish superficial and deep tissue dynamics opens up exciting possibilities for more accurate monitoring of cerebral blood flow in clinical settings such as brain-injured patients and brain-computer interfaces.

This research demonstrates a significant advancement in diffuse optical flowmetry with the introduction of interferometric near-infrared spectroscopy (iNIRS). The addition of optical phase and time-of-flight (TOF) resolution provides a more comprehensive and accurate way to assess blood flow dynamics in human tissues. Key findings include the confirmation of Brownian motion as the dominant mechanism of RBC displacement, the identification of ‘dynamical snake paths,’ and the ability to deconstruct and enhance classical DWS/DCS measurements. Future research directions include developing a more comprehensive theoretical framework to account for the complex light transport phenomena observed and exploring applications of iNIRS in diverse clinical settings.

While the study provides a significant advancement in the field, there are some limitations. The 5-parameter model used for fitting some data lacks a theoretical foundation, though the researchers provided an empirical justification. The study mainly focuses on blood flow dynamics, not accounting for other tissue properties that might affect light scattering. The sample size for some of the experiments is relatively small, which could limit the generalizability of the findings. Although the simulations were useful in interpreting results, there could be a mismatch between the simulations and real-world scenarios, depending on the accuracy of input parameters. Also, the study focused on a small selection of tissues and didn't generalize the findings to a broader range of tissue types and physiological states.