Direct measurement of tissue blood flow and metabolism with diffuse optics

Abstract

Diffuse optics has proven useful for quantitative assessment of tissue oxy- and deoxyhaemoglobin concentrations and, more recently, for measurement of microvascular blood flow. In this paper, we focus on the flow monitoring technique: diffuse correlation spectroscopy (DCS). Representative clinical and pre-clinical studies from our laboratory illustrate the potential of DCS. Validation of DCS blood flow indices in human brain and muscle is presented. Comparison of DCS with arterial spin-labelled MRI, xenon-CT and Doppler ultrasound shows good agreement (0.50<r<0.95) over a wide range of tissue types and source detector distances, corroborating the potential of the method to measure perfusion non-invasively and in vivo at the microvasculature level. All-optical measurements of cerebral oxygen metabolism in both rat brain, following middle cerebral artery occlusion, and human brain, during functional activation, are also described. In both situations, the use of combined DCS and diffuse optical spectroscopy/near-infrared spectroscopy to monitor changes in oxygen consumption by the tissue is demonstrated. Finally, recent results spanning from gene expression-induced angiogenic response to stroke care and cancer treatment monitoring are discussed. Collectively, the research illustrates the capability of DCS to quantitatively monitor perfusion from bench to bedside, providing results that match up both with literature findings and with similar experiments performed with other techniques.

Figure 1.

(a) Schematic of the DCS measurement. The light intensity detected at the output fibre fluctuates in time and is fed to an autocorrelator device that computes the normalized temporal intensity autocorrelation function (g₂(τ)) from photon arrival times. (b) Temporal intensity autocorrelation curves measured in the brain of a piglet (living and dead). (c) Examples of normalized temporal electric field autocorrelation curves (g₁(τ)) across a wide variety of biological animal and human tissues in their normal states, along with the Brownian (solid lines) model and random-flow (dashed-dotted lines) model best fits for 〈Δr²(τ)〉 (dots, raw data). The Brownian model fits the data better than the random-flow model (r refers to source–detector separation). (Online version in colour.)

Figure 2.

(a) Experimental schematic: blood was diverted to a pump and reservoir and then re-injected into the femoral artery. DCS data from the flow through the leg vasculature were acquired using a non-contact camera probe. (b) Temporal electric field autocorrelation curves obtained before (labelled as ‘normal’) and during the pump-perfused scenario. Blood was diluted by approximately 10 times in the pump-perfused measurement (modified from [1]). (Online version in colour.)

Figure 3.

(a) Experimental protocol showing the timing of the two flow measurement techniques and the clinical intervention. (b) Left: anatomical map showing the position of the optical probe on the head. Right: CBF map obtained from a xenon-CT scan after Xe infusion. (c) Correlation between CBF as measured by DCS and xenon-CT (modified from [54]). (Online version in colour.)

Figure 4.

(a) Temporal time course of rBF measured by DCS (left axis) and MRI (right axis) during a cuff occlusion experiment. (b) Quantitative correlation between the two techniques based on peak flow (modified from [29]). (Online version in colour.)

Figure 5.

(a) Schematic showing the experimental optical probe with two detectors and nine sources. The regions of interest (ROIs) on the brain are shown in the box, and they were defined relative to the bregma: ROI-1 is an ischaemic region, ROI-2 is a peri-ischaemic region towards midline, ROI-3 is a posterior peri-ischaemic region and ROI-4 is a contralateral control region. (b) Time traces of S_tO₂, CBF and CMRO₂ in the four ROIs; blue, black, green and red represent ROI-1, 2, 3 and 4, respectively; t=0 represents the beginning of the artery occlusion (modified from [19]). (Online version in colour.)

Figure 6.

(a) Schematic showing the experimental finger tapping activation set-up. (b) Data (n=5) with the probe located over the somatomotor cortex, and (c) 1 cm off-centre from the activation spot (modified from [40]). (Online version in colour.)

Figure 7.

(a) CBF changes in each hemisphere (left, blue bars; right, grey bars) after HOB angle manipulation in a healthy subject. CBF changes in a diseased population exhibited (b) impaired autoregulation in the injured hemisphere, but about 20% of the population presented a (c) paradoxical response, wherein CBF decreased at −5^°. (b,c) Blue bars, contralesional; grey bars, ipsilesional. (Online version in colour.)