Non-Invasive Monitoring of Microvascular Oxygenation and Reactive Hyperemia using Hybrid, Near-Infrared Diffuse Optical Spectroscopy for Critical Care

Summary

We describe a protocol to non-invasively and continuously measure absolute microvascular blood flow index and blood oxygen saturation using a multi-modal device based on near-infrared diffuse optics. We then evaluate the metabolic rate of oxygen consumption and reactive hyperemia utilizing a vascular occlusion test.

Abstract

The detection of levels of impairment in microvascular oxygen consumption and reactive hyperemia is vital in critical care. However, there are no practical means for a robust and quantitative evaluation. This paper describes a protocol to evaluate these impairments using a hybrid near-infrared diffuse optical device. The device contains modules for near-infrared time-resolved and diffuse correlation spectroscopies and pulse-oximetry. These modules allow the non-invasive, continuous, and real-time measurement of the absolute, microvascular blood/tissue oxygen saturation (StO₂) and the blood flow index (BFI) along with the peripheral arterial oxygen saturation (SpO₂). This device uses an integrated, computer-controlled tourniquet system to execute a standardized protocol with optical data acquisition from the brachioradialis muscle. The standardized vascular occlusion test (VOT) takes care of the variations in the occlusion duration and pressure reported in the literature, while the automation minimizes inter-operator differences. The protocol we describe focuses on a 3-min occlusion period but the details described in this paper can readily be adapted to other durations and cuff pressures, as well as other muscles. The inclusion of an extended baseline and post-occlusion recovery period measurement allows the quantification of the baseline values for all the parameters and the blood/tissue deoxygenation rate that corresponds to the metabolic rate of oxygen consumption. Once the cuff is released, we characterize the tissue reoxygenation rate, magnitude, and duration of the hyperemic response in BFI and StO₂. These latter parameters correspond to the quantification of the reactive hyperemia, which provides information about the endothelial function. Furthermore, the above-mentioned measurements of the absolute concentration of oxygenated and deoxygenated hemoglobin, BFI, the derived metabolic rate of oxygen consumption, StO₂, and SpO₂ provide a yet-to-be-explored rich data set that can exhibit disease severity, personalized therapeutics, and management interventions.

Introduction

Critically ill patients, particularly those with sepsis and other similar conditions, often exhibit impaired reactive hyperemia and microvascular oxygenation^1,2,3. During the first waves of COVID-19 pandemic, an unforeseen number of patients required intensive care management, during which the impact of the virus on the endothelium became evident but without a clear strategy to assess and manage^4,5,6. As a result, there has been a growing recognition of the importance of detecting endothelial dysfunction, which can be indirectly evaluated by reactive hyperemia, in critical care, i.e., the intensive care unit (ICU) populations⁷. A practical, robust, and widely available assessment of oxygen delivery and consumption to the tissues is expected to be of utmost importance in optimizing resuscitation strategies and directly addressing microcirculatory issues. Studies have consistently demonstrated that persistent microcirculatory alterations and lack of coherence between macrocirculation and microcirculation are, to some extent, predictive of organ failure and unfavorable outcomes in patients affected by septic shock or hemorrhagic shock, among other critical conditions, even when systemic parameters are considered to be normal^8,9,10. It has become evident that relying solely on macrocirculatory parameters is inadequate, as microcirculation plays a critical role in tissue oxygenation and organ function^11,12,13. This paper describes a protocol that uses a new multi-modal device based on near-infrared diffuse optical technologies that has been developed within an international consortium that focuses on ICU patients. The project, VASCOVID (https://vascovid.eu), was motivated by the COVID-19 pandemic to evaluate microvascular health in peripheral muscles in intensive care. We have designed a protocol using the developed VASCOVID device that aims to enhance our understanding of these parameters and how these parameters can be useful in managing critically ill patients with a much broader scope than COVID-19 patients.

Near infrared spectroscopy (NIRS) has been utilized to assess microcirculation non-invasively for decades in a broad range of clinical applications including, the ICU patients^14,15,16,17. It is important to note that the simplest application of NIRS, i.e., continuous wave NIRS (CW-NIRS), is implemented in widely used and clinically approved devices^17,18, used for measuring the absolute concentrations of oxy- (HbO) and deoxy-hemoglobin (HbR) to calculate the blood/tissue oxygen saturation (StO₂) of the microvasculature. While these devices have found niche uses in clinical management, such as during cardiac surgery, they have clear limitations due to the physics of photon propagation in tissues. This means that their accuracy, precision, and repeatability are questionable, hence, they are often utilized as trend monitors^19,20. Furthermore, their results are heavily influenced by superficial tissues such as the overlaying adipose and skin layers.

Time-resolved NIRS (TRS) employs short laser pulses in the picosecond range at multiple wavelengths to assess their delay and broadening after traversing through a tissue²¹. This allows TRS to separate the effects of absorption from scattering to obtain robust, accurate, and precise estimates, also allowing it to calculate the total hemoglobin concentration (HbT). Since TRS also resolves pathlengths, it can be utilized to better separate superficial signals from the deep signals of interest^18,21. This comes at the cost of complexity, price, and bulkiness. However, in recent years, TRS systems have come down in complexity and cost, resulting in more accessible and easier-to-use devices. This manuscript describes a device that uses a compact original equipment manufacturer (OEM) commercial TRS module^22,23.

Diffuse correlation spectroscopy (DCS) is another near-infrared technology that utilizes the temporal statistics of diffuse speckles to quantify the movement of light-scattering particles, which are dominated by red blood cells in tissues^16,24. This, in turn, is well known to be an indicator of microvascular blood flow, which we refer to as the blood flow index (BFI)²⁵. The simultaneous use of TRS and DCS in a hybrid optical device offers insights into oxygen metabolism by utilizing common models to derive the local oxygen extraction fraction and multiplying by the blood flow^15,26,27.

In order to assess the microcirculation at the ICU, NIRS is often utilized with a vascular occlusion test (VOT), which is an ischemic challenge that is performed by blocking the blood supply to the probed peripheral muscle for a certain duration (a few minutes)^{28,29,30,31,32}. Most commonly, it is executed by inflating a tourniquet wrapped around the upper arm above the systolic pressure³³. During the VOT, the clinicians assess the response of the microvascular blood oxygenation to changes in blood flow to derive oxygen metabolism at rest and reactive hyperemia³⁴. The assumption is that during the VOT, with the cuff inflated well above the limb occlusion pressure, there is no inflow or outflow of blood. Therefore, the start of VOT shows a downward slope of StO₂, i.e., deoxygenation (DeO₂), as oxygen is consumed by the tissue, which allows an estimate of the metabolic rate of oxygen consumption. When the VOT ends and the cuff is deflated, blood rushes in to compensate for its depletion, leading to a hyperemic response. This rush generates a sharp upward slope in StO_2, i.e., a reoxygenation (ReO₂). The hyperemic response, which is an increase beyond the initial baseline with a slow recovery back to the baseline, estimates the reactive hyperemia. The combination of NIRS with a VOT has gained increasing interest in intensive care due to its ease of use and potential for predicting adverse outcomes and even mortality in critical conditions such as sepsis^35,36,37.

During the COVID-19 pandemic, our groups have initiated a worldwide consortium and recently completed the so-called HEMOCOVID-19 trial, showing an association between peripheral microcirculatory alterations and severity of acute respiratory distress syndrome in COVID-19 patients⁶. This was supported by other works as well^7,38. All these studies were done with the above-mentioned CW-NIRS systems, hence suffering from their shortcomings. Furthermore, the execution of VOT was not standardized across different studies and is affected by various parameters like occlusion duration, tourniquet pressure, and operator-based variations^29,39,40. A literature review clearly shows that for VOT and NIRS to gain traction in the clinics, it is important to measure blood flow, have standardized protocols, and have a robust NIRS system¹¹. Therefore, we have proposed that by utilizing a more advanced form of NIRS (TRS), measuring blood flow, and standardizing the cuff control during VOT, a better discrimination of pathological conditions from healthy ones could be achieved. To that end, we have developed this hybrid diffuse optical device that integrates multiple modules encompassing two near-infrared diffuse optical modules of TRS and DCS, pulse oximetry, and an automated tourniquet. The pulse oximetry module provides the heart rate (HR), perfusion index, and percentage of arterial oxygen saturation (SpO₂). A fast tourniquet is used in the device, which is critical for performing VOT. The device comes with an optional accessory box that allows us to acquire additional information during the use for extended and continuous quality control, such as routine and practical measurement of the instrument response function (IRF) for TRS and the measurement on a tissue-mimicking phantom for evaluating longitudinal stability. The device is shown as being utilized in the ICU in Figure 1.

Figure 1: Bedside arrangement of the portable device in the ICU with the probes and cuff attached to the patient. Please click here to view a larger version of this figure.

The multimodal smart probe incorporates source and detector optical fibers for both TRS and DCS with optical filters inside the device that prevent interference between DCS and TRS. The source-detector separation used in this system is 25 mm. Additionally, the probe incorporates a capacitive touch sensor, providing a valuable safety feature to prevent laser hazards according to the laser safety standard (IEC 60601-2-22:2019)⁴¹. The laser safety system within the device ensures that the laser emission occurs only when the probe is in contact with the tissue. If detachment of the probe is detected, the lasers are immediately switched off, ensuring the safety of both patients and operators. Moreover, the probe is integrated with an accelerometer, load sensor, and light sensor for additional functionality and data collection purposes.

This paper describes the automated protocol where we probe the brachioradialis muscle simultaneously with a VOT using the developed device. The protocol timeline is shown in Figure 2. The protocol is completely automated, and no operator interventions are needed throughout its execution. By leveraging the capabilities of this novel device, we aim to gain valuable insights that let the physicians understand the physiopathology of peripheral oxygen consumption better and also assess the ratio of oxygen consumption and delivery thereby helping them to improve patient care comprehensively and efficiently.

Figure 2: Protocol timeline. The patient is at rest throughout the timeline with 0 mmHg pressure at initial baseline and recovery period. The VOT is performed with a tourniquet inflated to a pressure of 50 mmHg higher than the patient's systolic blood pressure. Please click here to view a larger version of this figure.

Protocol

The study was approved by the local ethics committee at Parc Tauli Hospital Universitari. Informed and signed consent was obtained from the patients or their next of kin. Absolute contraindications for entering the protocol were: clinical suspicion or echographic confirmation of venous thrombosis in the studied arm, other vascular or traumatic injuries in the studied arm, skin loss of integrity or lesions that could hinder the probe placement.

1. Device self-test

1. Switch on the device. The device starts with in-house developed software.
2. Turn the safety key to ON position, place the probe completely inside the instrument response function (IRF) box, and press the Reset button on the probe if it is glowing.
3. Press the OK button on the popup dialog box and wait until the device is ready.
   NOTE: The device performs self-tests to ensure stable working. The user is notified by a popup message when the device is ready.

2. Optional IRF and phantom measurement

1. Press OK when the device is ready.
2. Press Yes when it asks to measure an IRF. The device automatically adjusts the laser intensity to reach the desired count rate of 1 million.
3. Press the Stop button when a stable count rate and DTOF are observed. This IRF is saved in files as well as loaded in the software to be utilized for real time calculations.
4. Insert the probe in the Phantom box properly so that the probe attached indicator is on.
5. Press the Phantom button to start the phantom protocol.
   NOTE: The quality control test verifies that a sufficient number of photons are received by the DCS and TRS detectors and also checks if the dark counts are within the desired limits.
6. Continue recording for at least 30 s after the quality control to have a sufficient amount of data saved for further offline analysis.

3. Bedside measurement preparation

1. Attach the tourniquet on the upper arm above the elbow as done during a blood pressure measurement. Do not wrap the cuff loosely or very tightly around the arm.
   NOTE: Loosely attaching the tourniquet requires more air to reach the desired pressure. Slow inflation allows the body to readjust its physiology.
2. Attach the pulse oximeter to the index finger of the same arm. If it is not possible to attach to the index finger, attach it to any other finger.
3. Locate the brachioradialis muscle to be probed, which is in the lateral forearm just below the elbow. Ask the patient to open and close a fist to feel the muscle by placing fingers on the forearm. In the case of sedated patients or if they cannot move, trace the muscle by slightly twisting the arm with one hand. Feel the muscle between the thumb and fingers of the other hand.
4. Measure the arm circumference around the located muscle using a soft measuring tape, as shown in Figure 3.
5. Measure the approximate adipose tissue thickness on the top of the muscle by using a digital body fat caliper, as shown in Figure 4.
6. Attach the probe head to the muscle with the optical fibers and cables going toward the hand, as shown in Figure 5.
   NOTE: Do not attach the probe tightly; it can affect tissue physiology. Make sure the fibers are not touching moving objects, and it can create artifacts in the data.
7. Cover the probe with a black cloth to block the external light.
   NOTE: If the patient is awake, inform him that the VOT can cause a tingling sensation and he should not move the arm.

Figure 3: Measuring the arm circumference around the brachioradialis muscle. Please click here to view a larger version of this figure.

Figure 4: Measuring the adipose tissue thickness on top of the muscle using a body fat caliper. Please click here to view a larger version of this figure.

Figure 5: Probe attached to the muscle with fibers and cables going toward the hand. Please click here to view a larger version of this figure.

4. Data acquisition

1. Ensure that the probe-attached LED indicator on the device's front panel is shining and the touch icon in the software is green, showing that the probe is attached.
2. Press the Protocol timed button. Ensure it opens a new dialogue box, as shown in Figure 6. Enter the subject ID, operator ID, and the target pressure of 50 mmHg higher than the systolic blood pressure.
3. Press OK to start the automated protocol. Real-time data is displayed in the graphs. The protocol starts with quality control that automatically adjusts laser power and checks the photon counts and the interference between modalities. The quality check is completed within 2 min. Observe the circular icons labeled TRS and DCS, which must turn green at the end of the data quality check.
   NOTE: The green icons show that the photon count rate is within the desired range, there is no external light entering the probe, and there is no crosstalk between modalities. Hence, the measurement can be continued. The graphs are reset at the end of the quality phase, and signals representing patient data are plotted in real time.
4. Continue from step 2.6 if the TRS and DCS icons do not turn green and remain red at the end of the quality check. Press the Stop button to abort the protocol if the patient is unstable or requires sudden clinical intervention at any instant during the protocol.
5. Press the Extend button to add 30 s of pre-occlusion duration if the patient moves the arm and does not have stable baseline signals.
   NOTE: The operator can press the Extend button as many times and in any phase as required; each button press will add 30 s.
6. Ensure that the tourniquet automatically inflates to the desired pressure to start the VOT. Press + or - buttons to increase or decrease the desired occlusion pressure in steps of 5 mmHg if the blood pressure of the patient changes after starting the protocol. The start and stop of the VOT are automatically marked with yellow vertical lines.
   NOTE: The software is set to continuously acquire data and to automatically perform 3 min of VOT after 3 min of baseline. The pre-defined standard protocol lasts for six more minutes after the completion of VOT to evaluate the recovery after the patient's hyperemic response is over and a stable condition is obtained.
7. Press OK when the operator is notified at the completion of the protocol through a popup notification, which marks the successful completion of the protocol.
8. Remove the probes and the cuff from the patient and clean them using an alcohol swab or the equivalent.
9. Write down the clinical and demographic information (according to the pre-defined study protocols) along with the circumference of the arm at the probe location and the thickness of the overlying adipose tissue in the patient data form manually.

Figure 6: Screenshot of protocol parameters used for automatically executing the whole protocol. Please click here to view a larger version of this figure.

5. Data analysis

1. Use a script/program written in one's favorite language (example Python or MATLAB) to open and visualize recorded binary data.
2. Calculate the index of oxygen consumption that represents tissue metabolism and is defined as:
   
   where Hb is the hematocrit, which is recorded from clinical charts of the patient in the patient data form.
3. Compute the rate of DeO₂ (slope of StO₂ from the onset of VOT to 1 min), amplitude of DeO₂ (baseline StO₂- minimum StO₂), rate of ReO₂ (slope of StO₂ from completion of VOT to reaching peak value), amplitude of hyperemic peak of StO₂ and BFI (peak values), and area under the curve (AUC) of the reactive response after VOT for both StO₂ and BFI.
   NOTE: The computation of real-time absolute values of HbO, HbR, HbT, and StO₂ is achieved by fitting algorithm using the distribution of time of flight (DTOF) curves from TRS of both wavelengths. The theoretical details can be found in Torricelli et al. and Contini et al.^18,21. The calculation of BFI in real-time is achieved by the fitting algorithm using the autocorrelation curves from DCS. The theoretical details can be found in Durduran and Yodh¹⁶.

Discussion

We have demonstrated a fully automated, robust, non-invasive device for the continuous measurement and monitoring of skeletal muscle using hybrid diffuse optics for evaluating microvascular oxygenation, blood perfusion, and reactive hyperemia. Using this protocol with VASCOVID device, we can simultaneously measure absolute hemodynamic parameters of HbO, HbR, and HbT; oxygen saturation from StO₂ and SpO₂; DeO₂ and ReO₂; and BFI. The displayed real-time StO2 and BFI are obtained from the raw data of the previous second from the TRS and DCS modules, respectively. The fitting procedure is not time-consuming since modern processors use standard models of a semi-infinite, homogeneous medium. The parameters acquired do not paint the complete picture of endothelial function. However, reactive hyperemia measured has demonstrated prognostic value in several acute conditions where the endothelial impairment plays a major role, such as septic shock or COVID-19.^6,28 The protocol also incorporates an automated quality check that records the device parameters, which are useful for a research protocol in case an unexplained anomaly is detected later in any patient's data.

The quantification of the overlaying adipose layer and the arm circumference is important while measuring the brachioradialis muscle in this protocol since the photons pass primarily through the overlaying tissues when injected and when detected. It is well known in diffuse optics that there is an associated partial volume effect. Therefore, the superficial information should be recorded and utilized when analyzing the data in order to account for the effect of variations in adipose tissue^46,47. This is further amplified in these patient populations of interest since it is common in ICU patients to develop edema where the limbs are swollen as water is trapped due to immobilization and other reasons⁴⁸. In such patients, the variation in circumference during ICU stay can provide information about the severity of edema. The pathway of the light source reaching the detectors has to pass through all the superficial layers.

The cuff should be comfortably wrapped around the arm, ensuring a close fit. However, it is important to avoid excessive tightness that could exert excessive pressure on the arm solely through the act of wrapping the cuff⁴⁹. The goal is to achieve a secure and comfortable fit without causing unnecessary compression, which can alter baseline hemodynamic parameters. If it compresses the arm, the data quality will be compromised for the whole protocol, and the exerted pressure will be effectively added to the target pressure of VOT. In case the cuff is loosely wrapped to the arm, more air will be required to reach the target pressure and hence more time will be taken. This can give time to tissue for adjusting physiology as oxygen supply is being reduced slowly, which should be avoided⁵⁰.

It is important to attach the smart probe in a manner that maintains proper contact without exerting excessive pressure on the tissue. This allows for reliable measurements while avoiding the risk of local ischemia. Local ischemia occurs when blood flow to the area is restricted leading to compromised circulation and potentially corrupting the measurements⁵¹.

The capacitive touch sensor on the probe is used by the laser safety system to ensure that the laser shines only when the probe is attached to the tissue. If the patient has high hair density on the arm, the sensitivity of the touch sensor can be compromised. The application of a thin transparent double tape on the sensor side of the probe can effectively mitigate the touch sensor issue. When the probe is attached to the hairy arm together with this tape, it provides a reliable and stable touch signal. Predefined cuts of this tape are available for the smart probe with separation between light sources and detectors. The separation is essential to prevent the formation of a direct light channel between source and detector windows, which can affect the quality of the measurements. The use of transparent double tape serves as a practical solution to enhance the reliability of touch sensing in these circumstances. If the touch sensing is lost during the protocol, it turns off the lasers and the measurement is lost. The probe also has a load sensor which could, in the future, be utilized as a back-up safety measure.

If the patient moves their arm or a small clinical intervention disrupts the stability of acquired signals during the baseline phase, resulting in sharp peaks, it is advisable to utilize the extend feature. This feature allows for the acquisition of a stable baseline for three minutes, ensuring consistent and reliable signal measurement.

It is important to consider that the patient's blood pressure may undergo significant changes after initiating the protocol, which can impact the ability to reach the target pressure of 50 mmHg higher than the systolic blood pressure for the VOT. These fluctuations in blood pressure may be influenced by various factors, such as the patient's physiological response, medication effects, or other clinical conditions⁵². Therefore, the target pressure should be adjusted by pressing "+" or "-" buttons if necessary to ensure consistent administration of the VOT.

The typical execution of VOT has limitations due to operator variability, which is addressed in this protocol by having an automatic VOT. We are using the strategy to set the occlusion pressure of 50 mmHg above the systolic blood pressure level. This method stops the blood flow and has been reported in previous studies for performing the VOT^53,54. The individualized target pressure for VOT in this protocol helps in avoiding vasoconstriction that can happen by fixing a general target pressure for VOT. Pain caused by an unnecessarily high pressure can affect the measurement causing vasoconstriction, e.g. in a patient with systolic pressure of 120 mmHg and target pressure of 200 mmHg or 250 mmHg²⁹. We note that patients admitted to ICUs face an increased risk of thrombosis, primarily due to factors such as prolonged immobility, and sedation⁵⁵. This implies that to avoid risks, this protocol cannot be used in patients suffering from thrombosis or thrombophlebitis.

The application of this protocol can be useful in the ICU population where impaired reactive hyperemia is a common feature and can contribute to microvascular abnormalities^3,56. The parameters acquired in this protocol, without operator interventions during the measurement, have been previously used in the literature singularly or in a small combination for sepsis, cancer, stroke etc. to distinguish pathological conditions^1,11,15,31. Therefore, we believe that the combination of these relevant parameters are beneficial for several clinical applications. The data recorded by this protocol can help to select appropriate therapeutic strategies to improve vascular health⁵⁷. The valuable insights on tissue oxygenation and blood flow dynamics during occlusion and reperfusion allow us to assess the adequacy of blood supply to vital organs. It can help in identifying tissue hypoxia and guiding interventions to optimize organ perfusion⁵⁸. By using real-time information on microvascular oxygenation and reactive hyperemia, it assists as an additional tool in guiding hemodynamic management, fluid resuscitation, and vasopressor therapy^59,60. This ensures that interventions are tailored to individual patient needs, optimizing tissue oxygenation and perfusion^61,62. Furthermore, in mechanically ventilated patients, evolutive changes in microvascular oxygenation and blood flow within a spontaneous breathing trial can be of utmost importance when evaluating the cardiovascular tolerance to meet and overcome the increased metabolic burden derived from the work of breathing without assistance². On that behalf, a daily critical and challenging decision for the ICU patients on mechanical ventilation is the weaning process, which ends when the patient is considered able to breathe by himself, and the endotracheal tube is removed. The longitudinal application of this protocol can be used to evaluate the effectiveness of interventions, track disease progression, and guide treatment strategies.

Materials

Name	Company	Catalog Number	Comments
Alcohol swabs	No specific	N/A	For cleaning the probes and cuff after measurement
Black cloth	No specific	N/A	For blocking ambient light
Blood pressure monitor	OMRON	N/A	Hopital ICU equipment or off the shelf product
Body fat Calliper	Healifty	3257040-6108-1618385551	For measuring the fat layer
Examination gloves	No specific	N/A	To be used for interacting with patients
Kintex tape	No specific	N/A	For attaching the probe on arm
Koban wrap	No specific	N/A	For attaching the probe on arm
Measuring tape	YDM Industries	25-SB-30-150V3-19-1	For measuring the arm circumference
Scissors	No specific	N/A	for cutting tapes
VASCOVID precommercial prototype	VASCOVID consortium	N/A	Integrated at ICFO