Cerebrovascular Reactivity Measurement with Functional Near Infrared Spectroscopy

Summary

Presented here is a protocol for imaging and measurement of cerebrovascular reactivity in humans with functional Near Infrared Spectroscopy (fNIRS). fNIRS is a novel imaging modality that captures the concentration changes of hemoglobin species in the brain’s outermost cortex under specific stimuli.

Abstract

Cerebrovascular reactivity (CVR) is the capacity of blood vessels in the brain to alter cerebral blood flow (either with dilation or constriction) in response to chemical or physical stimuli. The amount of reactivity in the cerebral microvasculature depends on the integrity of the capacitance vasculature and is the primary function of endothelial cells. CVR is, therefore, an indicator of the microvasculature’s physiology and overall health. Imaging methods that can measure CVR are available but can be costly, and require magnetic resonance imaging centers and technical expertise. In this study, we used fNIRS technology to monitor changes of oxyhemoglobin (HbO) and deoxyhemoglobin (HbR) in the cerebral microvasculature to assess the CVR of 15 healthy controls (HC) in response to a vasoactive stimulus (inhaled 5% carbon dioxide or CO₂). Our results suggest that this is a promising imaging technology that offers a non-invasive, accurate, portable, and cost-effective method of mapping cortical CVR and associated microvasculature function, resulting from a traumatic brain injury or other conditions associated with cerebral microvasculopathy.

Introduction

Vascular health in the cerebral cortex can be measured via the vessels’ ability to constrict or dilate under varying physiological conditions. Measuring vascular reactivity can be useful in the diagnosis and management of neurological conditions associated with cerebral microvascular dysfunction, like dementia, traumatic brain injury (TBI) and even aging^1,2,3,4. Additionally, CVR can be used as a predictive and/or pharmacodynamic biomarker for neurological disorders such as Alzheimer's⁵ or TBI^6,7,8,9,10. Well-established imaging methods exist to study CVR in human and animal subjects. A typical method includes functional magnetic resonance imaging (fMRI) in conjunction with an exogenous or endogenous stimulus, such as hypercapnia¹¹, breath holding, or acetazolamide². Lu et al.^12,13 demonstrated that a simple gas delivery system coupled with MRI- Blood Oxygen Level Dependent (MRI-BOLD) imaging generates an accurate whole brain CVR maps.

Disruptions to the cerebral vasculature’s blood flow, volume, and metabolic rate of oxygen produces changes in the tissue concentrations of HbO and HbR. Tissue absorption of light at the near infrared range is sensitive to changes in the concentration of hemoglobin species, such as HbO and HbR. Therefore, measuring backscattered light over time can quantify changes of HbO and HbR concentration in the outermost cortex (approximately 2 cm)¹⁵, and can be used to assess temporal hemodynamic variations¹⁶ including cerebrovascular reactivity (CVR)¹⁷.

In our research paradigm, we employ the fNIRS instrument with continuous wave function. The device is composed of 4 sources and 10 detectors, which create 16 source-detector pairs (see Figure 1). The source-detector pairs are molded together onto a silicone strap that can easily be set over the forehead and held in place with self-adherent wrap. The device measures light intensity at 730 and 850 nm and has an acquisition frequency of 2 Hz. This system was selected because it is patient-friendly, easy to wear, and collects data from the prefrontal cortex, a brain region particularly vulnerable to TBI. Fortunately, most other fNIRS systems are compatible with our CVR acquisition technique, differing only in the cortical regions measured based on the brain area of interest.

While fMRI is considered the gold standard for functional brain imaging, fNIRS technology has unique advantages for assessing CVR in comparison to fMRI. The fNIRS imaging technique provides a high temporal resolution (with a granularity of milliseconds) and can quantify changes in both HbO and HbR concentration, while fMRI only measures changes in HbR^18,19,20. Moreover, fNIRS instruments are portable, economical, and easier to operate than fMRI. Finally, fNIRS technology better resolves subject motion, which is necessary given that vascular challenges like hypercapnia are often used in combination with cognitive or physical study tasks²¹.

In this paper, a hypercapnia challenge, combined with fNIRS technology is presented. We measured CVR values and studied the reproducibility of this method, hoping to offer a reliable alternative to fMRI CVR measures.

Protocol

The participants were recruited under an institutional review board approved protocol (ClinicalTrials.gov NCT01789164). The equipment described in the protocol is ethically approved by our institution.

1. Prepare the Materials used for the Hypercapnia Challenge (Figure 2)

1. Inflate a 200 L Douglas bag (Item #1) with a pre-mixed canister of medical-grade gas which is comprised of 5% carbon dioxide, 21% oxygen, and 74% nitrogen until full_.
2. Insert two diaphragms (Item #3) into the two-way non-rebreathing valve (Item #4) to safeguard that the gas will only flow in one direction. Attach one port of the three-way valve (Item #2) to the Douglas bag (Item #1) via the gas delivery tube (Item #5), and the other port to the two-way non-rebreathing valve (Item #4) via a second gas delivery tube (Item #5).
3. Fasten the mouthpiece (Item #6) to the connector (Item #7) and then fasten the connector to the two-way non-rebreathing valve (Item #4).
4. Insert the capnograph tubing (Item #8) into the hole in the connector (Item #7).
5. Attach the air-filter (Item #9) to the capnograph tubing (Item #8).
6. Screw the end of the plastic air-filter (Item #9) that isn’t connected to the capnograph tubing (Item #8) into the CO₂ (Item #10) monitor.
7. Connect the capnograh (Item #10) to a laptop with a cable. Open the data port reader software, select the corresponding USB port and start the data reading. Turn on the capnograph. Data will automatically be displayed on the computer screen.
8. Connect the fNIRS box to the computer with a USB cable. Connect the source-detector headband to the FNIRS box. Connect the power adapter to the fNIRS box and turn on the switch.

2. Procedures during the experiment

1. Ask the participant to sit on a chair and to make themselves comfortable while setting up the devices. Turn the fNIRS system on.
2. Place the source-detector headband on the patient’s forehead, over the underlying prefrontal cortex areas (dorsal and inferior frontal cortical areas)²¹.
3. Check that the source-detector headband is carefully positioned above the eyebrow and in the middle of the forehead. Place the lower detector row approximately 3.5 cm above the nasion or bridge of the nose where the indentation of the upper nose meets the forehead between the eyes.
4. Make sure the detectors are firmly adhered to the participant’s skin without chromophores (e.g., make-up) or hair interfering. No skin preparation is needed.
5. Under “Device Setting”, set the gain for detectors between 1 and 20. A higher gain will increase the sensitivity of the light detectors. The default value is 20. Set the “LED Current” between 5 mA and 20 mA. Larger values will result in brighter light and will increase the signal level generated by the detectors. The default value is 20 mA.
6. In the acquisition software, press “Start Current Experiment”. Sources will send light at 2 wavelengths and light signal intensity detected from each detector will be displayed in real time. In case of saturated (signal>4,000) or low signal (signal <1,000), adjust the contact between the source-detector headband and the skin or the parameters in step 2.3 and 2.4. The exact full procedure has been explained in Ayaz et al.²².
7. Direct the participant to inhale and exhale through their mouth at their normal breathing pace. Fasten a nose clip on the participant’s nose and remind them to continue breathing normally through their mouth, and to alert someone if they feel any discomfort or have any difficulty breathing.
8. Carefully insert the mouthpiece (item#6) into the participant’s mouth so that they can continue to breathe. For increased participant comfort during the procedure, ask the participant to support the non-rebreathing valve (Item #3) with their hand.
9. Press the “Baseline” button in associated software. It will measure and automatically record the light signal for the fNIRS baseline for 20 seconds (s).
10. Press “Record” before starting the experiment.
11. At the beginning of the experiment, start the clock, press “Manual Marker” and write on a paper the time displayed by the capnograph. Every minute turn the valve connected to the gas tubing to cycle between room air and room air mixed with 5% CO₂. Again, press “Manual Marker” and write on a paper the time displayed by the capnograph each time the inhaled gas mixture is changed (Figure 3).
    NOTE: Manually marking the time displayed on the capnograph is essential for future synchronization between fNIRS optical signals and capnograph’s EtCO₂ trace.
12. After 7 min, stop the fNIRS recording by clicking the “Stop” button. Allow 60 additional seconds of recording for the end-tidal CO₂ (EtCO₂) and save the EtCO₂ data as ASCII within the data reader software.
13. Notify the participant that the procedure is completed. Carefully take off the nose clip and withdraw the mouthpiece. Offer a tissue to the participant to absorb any accumulated saliva from the procedure.

3. Clean up Procedures

1. Dispose of the capnograph tubing (Item #8), filter (Item #9), mouthpiece (Item #6) and nose clip.
2. Cleanse the reusable equipment. Detach the two-way valve (Item #4) from the gas-delivery tube (Item #5) and the connector tube (Item #7) and extract the diaphragms (Item #3). Immerse the diaphragms (Item #3), connector tube, and two-way valve (Item #4) in a container full of a medical-grade detergent disinfectant that is phosphate-free and contains surfactants for 20 min. Dilute the detergent with distilled water in a ratio of 1:64.
3. Wash items #1,4,7 with distilled water then place them on top of a clean counter with a sterile material such as a chux pad underneath. Allow them to air-dry before re-use.
4. Empty the Douglas bag.

4. Data analysis

1. Signal processing using fNIRS data processing software
   NOTE: Signal processing is the first step of the data analysis. It is done using an fNIRS data processing software (e.g., fNIRSoft) in order to remove noise or artefact in the data due to patient movement. Only data from the acquisition software are needed for this analysis.
   1. In data processing software, click on “Load File” to select and then upload the acquired fNIRS data.
   2. Click on “Refine” and a pop-up window will appear. Select “Raw Data” and press “Next”.
      1. Click on both the median filtering and the sliding window motion artifact rejection (SMAR)²³ tools to recognize and delete both motion artifact and saturated channels from the raw signal. Press “Apply”.
      2. Click on “Low Pass Frequency” filter to discard pulse and breathing component (Hanning filter, n=20, cutoff=0.1Hz)^21,24,25,26. Press “Apply”.
      3. Click on “Detrend” to eliminate the slow temporal variation. Press “Apply”.
   3. Click on “OXI” to transform the light intensity into HbO and HbR concentrations. Click on “Save” and then select MATLAB as the save file format.
2. Signal processing with MATLAB
   NOTE: The second part of the analysis is done using MATLAB in order to correlate the fNIRS signal with the time shifted EtCO₂. Data from the previous step (4.1.5) and data from the capnograph (EtCO₂ trace, step 2.12) are needed for processing the data.
   1. Import the EtCO₂ trace from the capnograph in MATLAB as two columns (one for time and the second for EtCO₂ values). Shift the EtCO₂ time with pre-calibrated time to correct the delay from the sampling tubing time.
      NOTE: This is the time difference between one breath to the mouthpiece and the appearance of that breath on the CO₂ recording. In this set-up, it was 15 s.
   2. Use the first time point recorded from the capnograph at the beginning of the experiment, step 2.11 as the starting point (t=0). Convert the EtCO₂ into seconds.
   3. Import the oxy and deoxyhemoglobin data from step 4.1.3 into MATLAB.
   4. Calculate the physiological delay between EtCO₂ (measured in the mouth) and the fNIRS signal (measured in the brain) by finding the higher correlation coefficient between these two signals at varying time shifts. (see MATLAB script attached for step 4.2.3 to step 4.2.6). The time shift with the higher correlation coefficient is considered the optimal time.
   5. Shift the EtCO₂ time course by the optimal time (obtained in step 4.2.4). Keep the time points that have both fNIRS and EtCO₂ data. The two-time series should have the same length.
   6. Calculate CVR values for each channel, which is the solution of the linear equation between HbO (or HbR) and EtCO₂ using the Cholesky decomposition in MATLAB.

Results

fNIRS was performed with hypercapnia challenge on 15 healthy participants. Exclusion criteria were history of TBI, pre-existing disabling neurological or psychiatric disorders or pregnancy. The participants had a mean age of 37.7 ± 16 years (range 20-55) and 20% were female. As shown in a similar fMRI study²⁸, a 60 s inhalation of 5% CO₂ was accompanied by an increase in EtCO₂ pressure as measured by capnography. In our study, the EtCO₂ trace was accompanied b...

Discussion

We were able to measure CVR using fNIRS and a CO₂ gas inhalation technique in 15 healthy volunteers. The CVR value measured is the correlation between the acquired fNIRS signal and the EtCO₂. The challenge is to accurately align the temporal EtCO₂ trace with the fNIRS signal, in other words, to account for the time that it takes for blood to travel from the pulmonary vascular system to the heart and then to the cerebral vasculature. The inter-channel variability is low (30%) and shows a u...

Materials

Name	Company	Catalog Number	Comments
Blue cuff	22254	Vacumed
CO2-Air Gas Mixture Size 200	R012000 2003	Roberts Oxygen
Diaphragm (Size: medium, Type: spiral)	602021-2608	Hans Rudolph
Douglas bag (200-liters capacity)	500942	Harvard Apparatus
Gas delivery Tube	1011-108	Vacumed
Gas sampling Tube	T4305	QoSINA
Hydrophobic filter	9906-00	Philips Medical Systems
Male luer	11547	QoSINA
Mouth piece (Silicone, Model #9061)	602076	Hans Rudolph
Nose clip (Plastic foam, Model #9014)	201413	Hans Rudolph
Three-way valve (100% plastic)	CR1207	Hans Rudolph
Two-way non-breathing valve (22mm/ 15mm ID)	CR1480	Hans Rudolph