Near‐infrared diffuse correlation spectroscopy: the future of non‐invasive assessment of skeletal muscle oxygenation?

Near-infrared spectroscopy (NIRS) as a technology for assessment of tissue oxygenation was overlooked until the 1970s when development of the technique showed that the optical properties of near-infrared light in tissue can be correlated with tissue oxygenation. By evaluating absolute absorption and scatter coefficients of, and differentiating between, oxygenated haemoglobin (HbO2) and deoxyhaemoglobin (Hb), NIRS can provide a non-invasive continuous measure of tissue oxygenation status. NIRS is commonly used to assess peripheral oxygen saturation () in the microvasculature of skeletal muscle. As NIRS devices are widely available commercially and are relatively inexpensive, the technology is used frequently in both research and medical settings, and NIRS is also a useful tool in understanding the determinants of sports performance. The technology cannot, however, directly measure blood flow, because changes in oxygenation could be a result of fluctuations in red blood cell saturation or delivery. Diffuse correlation spectroscopy (DCS) is a regional microvascular blood flow assessment technique that functions on similar source–detector hardware to typical NIRS. DCS treats the scatter of emitted photons as a function of the motion of cells within a target volume, and so, with excellent temporal definition, can determine relative changes in microvascular blood perfusion index (BPI) (Bi et al. 2015). Compared to more conventional measures of microvascular perfusion (e.g. spin-labelled magnetic resonance imaging and positron emission tomography), DCS technology is inexpensive and portable, and as such is a novel tool offering bed-side microvascular blood flow assessment. However, commercially available DCS apparatus is limited and is an order of magnitude more expensive than NIRS oxygenation assessment technology (∼$10,000–$50,000 for NIRS, ∼>$150,000 for commercial combined NIRS–DCS technology); to date, most validation studies have utilized custom-built devices at great expense and therefore the technology has yet to become widely used. To avoid errors in DCS-derived BPI due to uncertainties in the coefficients of absorption and scattering of photons in tissue, DCS can be combined with a method of determining said coefficients of the tissue in question, e.g. frequency-domain NIRS (Irwin et al. 2011). The combination of standard NIRS with DCS enables continuous non-invasive assessment of the metabolic rate of O2, that is, by evaluating both blood flow and deoxy/haemoglobin saturation, in the same temporal and spatial volume. Only a few studies have validated DCS in working human skeletal muscle (e.g. Yu et al. 2007). In this issue of the Journal of Physiology, for the first time in humans, Tucker et al. (2019) undertook to validate the combination of DCS with typical continuous-wave NIRS technology against conventional O2 delivery and utilization measurements [Doppler-derived forearm blood flow (FBF) and venous blood gas analysis] in the working skeletal muscle. The exercise protocol consisted of 4 min of rhythmic handgrip exercise using a handgrip dynamometer (i.e. 2 s of isometric contraction, 2 s of relaxation). Additionally, to explore the efficacy of NIRS–DCS across a wider range of physiological conditions, NIRS–DCS-derived measures were compared to conventional Fick-derived measures with the exercising arm held in different positions: below and above the level of the heart. These positions induce different arterial perfusion pressures in the limb, thereby modifying the local pressure of perfusion into the resistance vasculature without greatly altering systemic mean arterial pressure. The main hypothesis was that combined NIRS–DCS would provide indices of skeletal muscle oxygenation close to those of established techniques, namely Doppler ultrasound and venous blood gas analysis, during changes in perfusion pressure and rhythmic handgrip exercise. From NIRS saturation measures and DCS BPI measures the relative metabolic rate of O2 (MRO2) could be calculated. During neutral position, significant decreases in were reflected by significant decreases in NIRS-derived . Additionally, Doppler-derived blood flow and DCS-derived BPI both increased at the onset of exercise. The group also found a very strong correlation coefficient between Doppler and DCS blood flow at four different intensities of handgrip exercise in a small subset (r = 0.96, P < 0.001) (n = 4). This suggests a strong relationship between Doppler and DCS flow results, 92% of the variance of one being accounted for by the other. The group claim that these results suggest good concurrent validity between conventional methods and a combined NIRS–DCS method. Tucker et al. claim that an additional noteworthy finding is the discrimination of myogenic autoregulation via differences in bulk conduit blood delivery, microvascular perfusion and HbO2 desaturation made possible by NIRS–DCS comparison with Fick-derived measures. A delayed time to steady-state perfusion and enhanced desaturation indicate a myogenic response that maintains O2 delivery and increases extraction to preserve sufficient consumption under metabolic demands. A small number of practical considerations are of note that may have supplemented the strength of this validation study: confounding movement artefacts, assumed and values, use of flexor digitorum profundus and, as mentioned above, choice of statistical analyses. Some but not all of these were addressed in the study. A significant limitation to the application of both NIRS and DCS technologies is the impact of movement artefacts. As photon scatter is greatly affected by the movement of substances in a sample volume (in fact the basis of DCS), all diffuse optical spectroscopy (DOS) techniques are vulnerable to fluctuations due to movement of unsecured probes or subcutaneous tissue translation. To counter this, Tucker et al. secured the device probes with hook and loop straps, and NIRS–DCS data were recorded only during the relaxation phase of the work duty cycle. The latter tactic was well advised; however, the additional use of the clinical adhesive collodion may have increased the signal-to-noise ratio and reduced motion artefacts between probe and skin better than hook and loop straps (Yucel et al., 2014). While FDP is a deep muscle of the forearm, superficial access is feasible at a relatively small area just distal to the common flexor tendon. However, as this area is difficult to confidently locate with palpation alone, it would have been advantageous for Tucker et al. to locate it using ultrasound, as they did to locate deep veins. DOS probes report chromophores between source and detector, so any shielding of the FDP under superficial wrist flexors would impact NIRS–DCS results, which is a concern given the small area. Furthermore, FDP is not the most taxed muscle during typical handgrip dynamometry, flexor digitorum superficialis and the intrinsic hand muscles working hardest. Calculating with assumed at 100 mmHg, and assessed via fingertip pulse oximeter is not ideal as accuracy of blood gas measurements is crucial. Pulse oximetry provides , which requires an accurate in order to calculate , and such a calculation was not reported. Additionally assessing arterial perfusion pressure via servo-controlled finger photoplethysmography on the exercising hand means perfusion pressure readings would be confounded by intermittent occlusion through contraction. Arterial catheterization of the exercising arm radial artery, although being more invasive, would resolve both of these issues. Finally, Tucker et al.’s statistical analysis concerning concurrence between NIRS–DCS and Fick-derived measures would be improved and further supported by the use of Bland–Altman plots, which would reveal bias or consistent error in either technique. Pearson's correlations do not indicate whether there is truly concordance between measurement techniques, only that there is a relationship between the results of the compared techniques. Much of the study's conclusions were interpreted from similarities in curves fitted to time series data. Bland–Altman plots would better quantify and illustrate the validity of NIRS–DCS against Doppler-derived blood flow (Figure 1), and venous blood gas analysis. Teasing apart these differences in delivery, perfusion and extraction in skeletal muscle during exercise under different perfusion pressures was enabled by the combination of DCS with conventional continuous-wave NIRS. Coupling NIRS with DCS offers useful discrimination of temporal variations of tissue oxygenation in conjunction with fine fluctuations in microvascular perfusion. NIRS–DCS may have future clinical applications to evaluate skeletal muscle myogenic responsiveness in degenerative muscle pathologies and thereby aid in determining appropriate medicinal and training modalities. For example, gauging myogenic responsiveness may be useful in treating peripheral vascular diseases, where ischaemia due to arterial embolism or vasospasm are common, and the capability of the muscle to maintain perfusion and increase O2 extraction may be paramount. Evaluation of age-related microvascular dysfunction could also be enhanced with the use of NIRS–DCS. None. Sole author. None. Dr Philip N. Ainslie, Christina D. Bruce, MSc, and Hannah G. Caldwell, BHK, are thanked for their support, criticism, and invaluable discussion.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.