Diffuse Correlation Spectroscopy Analysis Implemented on a Field Programmable Gate Array

A Device-on-Chip Solution for Real-Time Diffuse Correlation Spectroscopy Using FPGA

Diffuse correlation spectroscopy (DCS) is a non-invasive technology for the evaluation of blood perfusion in deep tissue. However, it requires high computational resources for data analysis, which poses challenges in its implementation for real-time applications. To address the unmet need, we developed a novel device-on-chip solution that fully integrates all the necessary computational components needed for DCS. It takes the output of a photon detector and determines the blood flow index (BFI). It is implemented on a field-programmable gate array (FPGA) chip including a multi-tau correlator for the calculation of the temporal light intensity autocorrelation function and a DCS analyzer to perform the curve fitting operation that derives the BFI at a rate of 6000 BFIs/s. The FPGA DCS system was evaluated against a lab-standard DCS system for both phantom and cuff ischemia studies. The results indicate that the autocorrelation of the light correlation and BFI from both the FPGA DCS and the reference DCS matched well. Furthermore, the FPGA DCS system was able to achieve a measurement rate of 50 Hz and resolve pulsatile blood flow. This can significantly lower the cost and footprint of the computational components of DCS and pave the way for portable, real-time DCS systems.

DCS blood flow index underestimates skeletal muscle perfusion in vivo: rationale and early evidence for the NIRS-DCS perfusion index

Significance

Diffuse correlation spectroscopy (DCS) permits non-invasive assessment of skeletal muscle blood flow but may misestimate changes in muscle perfusion.

Aim

We aimed to highlight recent evidence that DCS blood flow index (BFI) misestimates changes in muscle blood flow during physiological perturbation and to introduce a novel approach that adjusts BFI for estimated changes in vasodilation.

Approach

We measured changes in muscle BFI during quadriceps and forearm exercises using DCS, the latter of which were adjusted for estimated changes in microvascular flow area and then compared to Doppler ultrasound in the brachial artery. Then, we compared adjusted BFI- and arterial spin labeling (ASL) MRI measures of gastrocnemius blood flow during reactive hyperemia and plantar flexion exercise.

Results

We observed little-to-no change in quadriceps BFI during maximal-effort exercise. Similarly, forearm BFI was modestly increased during handgrip exercise, but the magnitude was significantly lower than measured by Doppler ultrasound in the brachial artery. However, this difference was ameliorated after adjusting BFI for estimated changes in microvascular flow area. Similar observations were also observed in the gastrocnemius when directly comparing the adjusted BFI values to ASL-MRI.

Conclusions

Adjusting BFI for estimated changes in microvascular flow area may improve DCS estimates of muscle blood flow, but further study is needed to validate these methods moving forward.

Epinephrine iontophoresis attenuates changes in skin blood flow and abolishes cutaneous contamination of near-infrared diffuse correlation spectroscopy estimations of muscle perfusion

Near-infrared diffuse correlation spectroscopy (NIR-DCS) is an optical imaging technique for measuring relative changes in skeletal muscle microvascular perfusion (i.e., fold change above baseline) during reactive hyperemia testing and exercise and is reported as a blood flow index (BFI). Although it is generally accepted that changes in BFI are primarily driven by changes in muscle perfusion, it is well known that large, hyperthermia-induced changes in cutaneous blood flow can uncouple this relationship. What remains unknown, is how much of an impact that changes in cutaneous perfusion have on NIR-DCS BFI and estimates of skeletal muscle perfusion under thermoneutral conditions, where changes in cutaneous blood flow are assumed to be relatively low. We therefore used epinephrine iontophoresis to pharmacologically block changes in cutaneous perfusion throughout a battery of experimental procedures. The data show that 1) epinephrine iontophoresis attenuates changes in cutaneous perfusion for up to 4-h posttreatment, even in the face of significant neural and local stimuli, 2) under thermoneutral conditions, cutaneous perfusion does not significantly impact NIR-DCS BFI during reactive hyperemia testing or moderate-intensity exercise, and 3) during passive whole body heat stress, when cutaneous vasodilation is pronounced, epinephrine iontophoresis preserves NIR-DCS measures of skeletal muscle BFI during moderate-intensity exercise. Collectively, these data suggest that cutaneous perfusion is unlikely to have a major impact on NIR-DCS estimates of skeletal muscle BFI under thermoneutral conditions, but that epinephrine iontophoresis can be used to abolish cutaneous contamination of the NIR-DCS BFI signal during studies where skin blood flow may be elevated but skeletal muscle perfusion is of specific interest.

Lossless Compressed Sensing of Photon Counts for Fast Diffuse Correlation Spectroscopy

Diffuse Correlation Spectroscopy (DCS), a noninvasive optical technique, measures deep tissue blood flow using avalanche photon counting modules and data acquisition devices such as FPGAs or correlator boards. Conventional DCS instruments use in-processor counter modules that consume 32 bits/channel which is inefficient for low-photon budget situations prevalent in diffuse optics. Scaling these photon counters for large-scale imaging applications is difficult due to bandwidth and processing time considerations. Here, we introduce a new, lossless compressed sensing approach for fast and efficient detection of photon counts. The compressed DCS method uses an array of binary-coded-decimal counters to record photon counts from 8 channels simultaneously as a single 32-bit number. We validate the compressed DCS approach by comparisons with conventional DCS in experiments on tissue simulating phantoms and in-vivo arm cuff occlusion. Lossless compressed DCS was implemented with 87.5% compression efficiency. In tissue simulating phantoms, it was able to accurately estimate a tissue blood flow index, with no statistically significant difference compared to conventional DCS. Compressed DCS also recorded blood flow in vivo, in human forearm, with signal-to-noise ratio and dynamic range comparable to conventional DCS. Lossless 87.5% efficient compressed sensing counting of photon counts meets and exceeds benchmarks set by conventional DCS systems, offering a low-cost alternative for fast (~100 Hz) deep tissue blood flow measurement with optics.

Blood flow estimation via numerical integration of temporal autocorrelation function in diffuse correlation spectroscopy

BACKGROUND AND OBJECTIVE

Diffuse correlation spectroscopy (DCS) is an optical technique widely used to monitor blood flow. Recently, efforts have been made to derive new signal processing methods to minimize the systems used and shorten the signal processing time. Herein, we propose alternative approaches to obtain blood flow information via DCS by numerically integrating the temporal autocorrelation curves.

METHODS

We use the following methods: the inverse of K² (IK2)-based on the framework of diffuse speckle contrast analysis-and the inverse of the numerical integration of squared g₁ (INISg1) which, based on the normalized electric field autocorrelation curve, is more simplified than IK2. In addition, g₁ thresholding is introduced to further reduce computational time and make the suggested methods comparable to the conventional nonlinear fitting approach. To validate the feasibility of the suggested methods, studies using simulation, liquid phantom, and in vivo settings were performed. In the meantime, the suggested methods were implemented and tested on three types of Arduino (Arduino Due, Arduino Nano 33 BLE Sense, and Portenta H7) to demonstrate the possibility of miniaturizing the DCS systems using microcotrollers for signal processing.

RESULTS

The simulation and experimental results confirm that both IK2 and INISg1 are sufficiently relevant to capture the changes in blood flow information. More interestingly, when g₁ thresholding was applied, our results showed that INISg1 outperformed IK2. It was further confirmed that INISg1 with g₁ thresholding implemented on a PC and Portenta H7, an advanced Arduino board, performed faster than did the deep learning-based, state-of-the-art processing method.

CONCLUSION

Our findings strongly indicate that INISg1 with g₁ thresholding could be an alternative approach to derive relative blood flow information via DCS, which may contribute to the simplification of DCS methodologies.

FPGA Correlator for Applications in Embedded Smart Devices

Correlation has a variety of applications that require signal processing. However, it is computationally intensive, and software correlators require high-performance processors for real-time data analysis. This is a challenge for embedded devices because of the limitation of computing resources. Hardware correlators that use Field Programmable Gate Array (FPGA) technology can significantly boost computational power and bridge the gap between the need for high-performance computing and the limited processing power available in embedded devices. This paper presents a detailed FPGA-based correlator design at the register level along with the open-source Very High-Speed Integrated Circuit Hardware Description Language (VHDL) code. It includes base modules for linear and multi-tau correlators of varying sizes. Every module implements a simple and unified data interface for easy integration with standard and publicly available FPGA modules. Eighty-lag linear and multi-tau correlators were built for validation of the design. Three input data sets-constant signal, pulse signal, and sine signal-were used to test the accuracy of the correlators. The results from the FPGA correlators were compared against the outputs of equivalent software correlators and validated with the corresponding theoretical values. The FPGA correlators returned results identical to those from the software references for all tested data sets and were proven to be equivalent to their software counterparts. Their computation speed is at least 85,000 times faster than the software correlators running on a Xilinx MicroBlaze processor. The FPGA correlator can be easily implemented, especially on System on a Chip (SoC) integrated circuits that have processor cores and FPGA fabric. It is the ideal component for device-on-chip solutions in biosensing.

Fast diffuse correlation spectroscopy with a low-cost, fiber-less embedded diode laser

Diffuse correlation spectroscopy (DCS), a popular optical technique for fast noninvasive measurement of blood flow, is commonly implemented using expensive fiber-coupled long coherence length laser systems. Here, we report the development of a portable and fiber-less approach that can be used as a low-cost alternative to illuminate tissue in DCS instruments. We validate the accuracy and noise characteristics of the fiber-less DCS laser source, by comparisons against traditional DCS light sources, with experiments on controlled tissue-simulating phantoms and in humans.

Towards rapid intraoperative axial localization of spinal cord ischemia with epidural diffuse correlation monitoring

Spinal cord ischemia leads to iatrogenic injury in multiple surgical fields, and the ability to immediately identify onset and anatomic origin of ischemia is critical to its management. Current clinical monitoring, however, does not directly measure spinal cord blood flow, resulting in poor sensitivity/specificity, delayed alerts, and delayed intervention. We have developed an epidural device employing diffuse correlation spectroscopy (DCS) to monitor spinal cord ischemia continuously at multiple positions. We investigate the ability of this device to localize spinal cord ischemia in a porcine model and validate DCS versus Laser Doppler Flowmetry (LDF). Specifically, we demonstrate continuous (>0.1Hz) spatially resolved (3 locations) monitoring of spinal cord blood flow in a purely ischemic model with an epidural DCS probe. Changes in blood flow measured by DCS and LDF were highly correlated (r = 0.83). Spinal cord blood flow measured by DCS caudal to aortic occlusion decreased 62%. This monitor demonstrated a sensitivity of 0.87 and specificity of 0.91 for detection of a 25% decrease in flow. This technology may enable early identification and critically important localization of spinal cord ischemia.

Impact of changes in tissue optical properties on near-infrared diffuse correlation spectroscopy measures of skeletal muscle blood flow

Near-infrared diffuse correlation spectroscopy (DCS) is increasingly used to study relative changes in skeletal muscle blood flow. However, most diffuse correlation spectrometers assume that tissue optical properties-such as absorption (μ_a) and reduced scattering (μ'_s) coefficients-remain constant during physiological provocations, which is untrue for skeletal muscle. Here, we interrogate how changes in tissue μ_a and μ'_s affect DCS calculations of blood flow index (BFI). We recalculated BFI using raw autocorrelation curves and μ_a/μ'_s values recorded during a reactive hyperemia protocol in 16 healthy young individuals. First, we show that incorrectly assuming baseline μ_a and μ'_s substantially affects peak BFI and BFI slope when expressed in absolute terms (cm²/s, P < 0.01), but these differences are abolished when expressed in relative terms (% baseline). Next, to evaluate the impact of physiologic changes in μ_a and μ'_s, we compared peak BFI and BFI slope when μ_a and μ'_s were held constant throughout the reactive hyperemia protocol versus integrated from a 3-s rolling average. Regardless of approach, group means for peak BFI and BFI slope did not differ. Group means for peak BFI and BFI slope were also similar following ad absurdum analyses, where we simulated supraphysiologic changes in μ_a/μ'_s. In both cases, however, we identified individual cases where peak BFI and BFI slope were indeed affected, with this result being driven by relative changes in μ_a over μ'_s. Overall, these results provide support for past reports in which μ_a/μ'_s were held constant but also advocate for real-time incorporation of μ_a and μ'_s moving forward.NEW & NOTEWORTHY We investigated how changes in tissue optical properties affect near-infrared diffuse correlation spectroscopy (NIR-DCS)-derived indices of skeletal muscle blood flow (BFI) during physiological provocation. Although accounting for changes in tissue optical properties has little impact on BFI on a group level, individual BFI calculations are indeed impacted by changes in tissue optical properties. NIR-DCS calculations of BFI should therefore account for real-time, physiologically induced changes in tissue optical properties whenever possible.

Multi-Site Optical Monitoring of Spinal Cord Ischemia during Spine Distraction

Optimal surgical management of spine trauma will restore blood flow to the ischemic spinal cord. However, spine stabilization may also further exacerbate injury by inducing ischemia. Current electrophysiological technology is not capable of detecting acute changes in spinal cord blood flow or localizing ischemia. Further, alerts are delayed and unreliable. We developed an epidural optical device capable of directly measuring and immediately detecting changes in spinal cord blood flow using diffuse correlation spectroscopy (DCS). Herein we test the hypothesis that our device can continuously monitor blood flow during spine distraction. Additionally, we demonstrate the ability of our device to monitor multiple sites along the spinal cord and axially resolve changes in spinal cord blood flow. DCS-measured blood flow in the spinal cord was monitored at up to three spatial locations (cranial to, at, and caudal to the distraction site) during surgical distraction in a sheep model. Distraction was halted at 50% of baseline blood flow at the distraction site. We were able to monitor blood flow with DCS in multiple regions of the spinal cord simultaneously at ∼1 Hz. The distraction site had a greater decrement in flow than sites cranial to the injury (median -40 vs. -7%,). This pilot study demonstrated high temporal resolution and the capacity to axially resolve changes in spinal cord blood flow at and remote from the site of distraction. These early results suggest that this technology may assist in the surgical management of spine trauma and in corrective surgery of the spine.