mNIRS

Muscle near infrared spectroscopy (mNIRS) has developed enormously since Frans Jöbsis first demonstrated in 1977 that near infrared light could be used to monitor oxygen-sensitive signals in living tissue. Today there is a wide range of instruments and an equally wide range of ways in which NIRS is used to study muscle. This is a strength of the field, but it can also make it difficult to understand exactly what different instruments measure, how those measurements are obtained and, most importantly, what they tell us about muscle physiology. I’ve been working with NIRS and oxygen metabolism for many years, and I’ve created MuscleNIRS as a place to discuss some of these questions. The site has two main parts. The  Education  section contains short introductions to the measurements that can be made with mNIRS and the methods used to make them. These are intended to be relatively permanent resources and will be expanded over time. The  Blog  will be less formal and mor...

Methods other than SRS use multiple source–detector distances to estimate StO₂. While the methods vary, a shared principle is that shorter distances sample more superficial tissues, while longer distances sample deeper regions — such as the cerebral cortex or muscle tissue. These methods are sometimes loosely referred to as spatial resolution or self-calibrating approaches and, confusingly, the term spatially resolved spectroscopy is also occasionally used. Importantly, however, these methods do not use the same analytical approach as the classic SRS algorithms used in systems such as the Hamamatsu NIRO or Artinis PortaMon. Among commercial devices, the most widely used has probably been the INVOS system, now sold by Medtronic. The traditional adult sensor uses detectors placed 3 cm and 4 cm from the source. The signal from the shorter separation, which is more sensitive to superficial tissue, is used to reduce the superficial contribution to the signal measured at the longer separatio...

There are methods for measuring absolute muscle tissue oxygen saturation (StO₂) that do not require expensive time-domain or frequency-domain systems. One such method is Spatially Resolved Spectroscopy (SRS). Originally developed in the Hamamatsu NIRO systems, SRS is now best known for its implementation in devices such as the Artinis PortaMon. SRS assumes homogeneous scattering within tissue. To visualize this, imagine shining a torch into a dilute scattering solution such as milk: the light that emerges appears broadly diffused, and its brightness does not change much with small changes in viewing position. This is because scattering redirects light in many directions, creating a relatively uniform glow. If you add a chromophore to the milk, the brightness decreases with distance because of absorption. In an SRS measurement, a single light source is combined with multiple detectors placed close together relative to their distance from the source (or vice versa). The slope of the chan...

This simple approach can be applied to almost any mNIRS system. The principle is to expose the muscle to physiological conditions that produce a minimum and maximum oxygenation signal, and use these to define a 0–100% calibration range. The minimum can be produced by exercise, arterial occlusion using a pressure cuff, or a combination of the two. Exercise increases muscle oxygen consumption, whilst arterial occlusion prevents oxygen delivery, allowing tissue oxygenation to fall towards a minimum. The upper end of the calibration can also be defined in different ways. A commonly used approach is to release an arterial occlusion and use the peak oxygenation produced by the resulting reactive hyperemia as the 100% value. Earlier approaches have also used other physiological reference points. The resulting 0–100% scale is therefore a physiological calibration rather than an absolute measurement of oxygen saturation. Exactly what the 0% and 100% points represent depends upon the calibration...

 Whilst all mNIRS devices can in principle measure changes in oxy and deoxy heme concentrations, more expensive equipment is needed to measure absolute concentrations. NIR light is multiply scattered in tissue. Light reaching the detector is therefore a function of both the tissue absorption and scattering properties. A time-domain system measures the time at which photons arrive at the detector. A short laser pulse is detected as a spread of photons, the so-called temporal point spread function (TPSF). The beauty of the TPSF is that its shape contains information that can be used to separate absorption and scattering. Essentially, the earliest photons have travelled shorter paths through the tissue, whilst later photons have travelled progressively longer paths. The shape of the TPSF can therefore be fitted to a model of light diffusion to obtain an absorption coefficient (μa) and reduced scattering coefficient (μs′) at each wavelength. Whilst scattering information may be interes...

 Early mNIRS measurements were reported in “arbitrary units” or used a “physiological calibration” (see later method). David Delpy and colleagues, together with Hamamatsu Photonics, pioneered measurements of changes in heme concentration in “real” units, enabling subsequent quantitative measurements of variables such as blood flow and oxygen consumption (see later methods). The Beer-Lambert Law states that changes in absorbance are directly proportional to three things: the change in chromophore concentration; an intrinsic property of the chromophore at the measuring wavelength (the extinction coefficient); and the optical pathlength. With known extinction coefficients, it is then straightforward to calculate concentration changes if you use the same or a greater number of optical wavelengths than chromophores. For example, the Hamamatsu NIRO-500 had four measuring wavelengths and measured three unknowns — oxyhemoglobin, deoxyhemoglobin and cytochrome oxidase — in units of µM·cm. T...

Near infrared light (700–1,000 nm) can penetrate tissue because of the relatively low absorbance of biological chromophores in this wavelength range. However, those chromophores that do absorb NIR light can change colour in response to changes in oxygen supply or metabolism. In particular, oxyhemoglobin, oxymyoglobin, deoxyhemoglobin, deoxymyoglobin and mitochondrial cytochrome c oxidase have significant oxygen-sensitive signals. In general, myoglobin and hemoglobin cannot be distinguished spectrally because of their structural similarity (although see a later method). Therefore, there are essentially three NIR-detectable chromophores:  oxy[heme] ,  deoxy[heme]  and mitochondrial cytochrome c oxidase. These were first reported in the seminal 1977 paper by Frans Jöbsis: Jöbsis FF.   “Noninvasive, infrared monitoring of cerebral and myocardial oxygen sufficiency and circulatory parameters.”   Science  198 (1977): 1264–1267. This paper showed the anoxic–normox...