**1. Introduction**

Near-infrared spectroscopy (NIRS) is a spectroscopic technique which uses the NIR region of the electromagnetic spectrum to gain information about natural samples through their absorption of NIR light. This method is used in several branches of science. In medicine, it was first used in adult patients, who were placed on by-pass during cardiac surgery to follow cerebral oxygenation, cerebral rSO2 (rSO2-c,) and thereby perfusion and metabolism of the brain. Its many other possibilities soon became apparent. Although the brain remains the main organ of interest in patients of all ages, other tissues are being studied as well. Aside from cardiac surgery clinicians in specialties such as sports medicine, plastic surgery (to assess flap viability), and neonatology apply NIRS in clinical settings. (Feng et al., 2001)

By the late 1980's the first studies on monitoring of regional oxygenation in the neonatal brain were published. (Delpy et al., 1987; Edwards et al., 1988) In 2004 on average one new article on NIRS was published in Pub Med every day. (Ferrari et at, 2004) Monitoring of vital signs in the ICUs has scientific and patient care related goals. One may be able to gain better understanding of physiology and be alerted to changes in patient status to be able to respond immediately.

The vulnerability of the neonate, especially of the newborn brain, to changes in oxygenation is an ever present concern as it is linked to long-term outcome. For that reason neonatologists are obligated to find ways to monitor their patients to be ahead of evolving pathology and avoid the severe impact of negative events.

As early as 1999 the NINDS and NIH hosted a workshop for experts in the fields of neurology and neonatology to discuss the use of NIRS for cerebral monitoring in infants. The panel determined that the best NIRS instrument should be selected and used in longitudinal, blinded studies. Obtained data would need to be compared with short term, intermediate and long term outcomes. The questions the panel suggested to investigate were the predictive value of NIRS and its usefulness in leading to timely interventions and prevention of long term injury. (www.ninds.nih.gov/news\_andevents/proceedings/

Use of Near-Infrared Spectroscopy in the Management of Patients in

or tissue oxygenation index (TOI).

al., 2006)

**1.3 Validation** 

Neonatal Intensive Care Units – An Example of Implementation of a New Technology 7

NIRS monitors with somewhat different technology and algorithms available commercially (Wolf & Greisen, 2009) to measure the venous weighted regional oxygen saturation (rSO2)

Due to the small size and the thin covering layers of tissue of both term and preterm neonates, r-SO2/TOI measurements at a depth of 2-3 cm can reach brain, kidney, gut/ splanchnic circulation, liver and muscle. The access to these critical organs promises valuable physiologic information through monitoring by NIRS. Measurements of several sites can be recorded simultaneously. (Hoffman et al., 2003; McNeill et al., 2010, 2011)

NIRS measurements are organ specific and regional (rSO2), reflecting perfusion and metabolism by non-invasive measurement in real-time. They are not temperature, pulsatility or flow dependent. Thus they may offer advantages over traditional measures of perfusion such as capillary refill, blood pressure, and urine output, lactate, venous and arterial O2 which tend to alert the clinician once the disease process is further progressed. R-SO2 measurements cannot stand alone. While they may often be the first sign of change, they need to be interpreted in the context of other measurements such as mean arterial blood pressure (MABP), pulse oximetry (O2sat), blood gases, additionally in the research setting with measurements of cerebral blood flow (CBF) and cerebral blood volume (CBV). Evaluation of the link between the venous weighted NIRS readings and peripheral pulse oximetry, a measure of arterial O2, gives insight into oxygen supply and demand. Using a simple equation, the fractional extraction of oxygen (FTOE = SaO2-rSO2/SaO2) oxygen consumption can be calculated and oxygen supply can be assessed. (Lemmers et

NIRS was implemented by many enthusiastic clinicians without a vast body of previous research evidence. This phenomenon may be representative of an era of limited funding for larger studies linked with the promise of a non-invasive "safe" monitoring technology.

Before human application the initial research applying NIRS to measure rSO2 technology in the medical field occurred in the laboratory: One of the first examples of validation used a phantom brain model in which O2, N2, and CO2 content of a blood perfusate could be altered during measurements. The results correlated with findings in animal models. (Kurth et al., 1995) Later NIRS was further validated for the neonatologist in a newborn piglet model. The carotid, renal and mesenteric arteries were occluded and reperfused. These interventions led to rapid, simultaneous changes in rSO2 of the affected end-organs. (Wider, 2009) Furthermore, there have been validations in patients during intensive care, extracorporeal membrane oxygenation (ECMO) and cardiac surgery by comparing central blood samples with NIRS values. (Abdul-Khaliq et al., 2002; Benni et al., 2005; Nagdyman et al., 2004; Rais-Bahrami K et al, 2006; Weiss, 2005) Menke found reproducibility to be good as well. (Menke et al., 2003). The accuracy of data is impacted by light scattering, hemoglobin concentration and chromophores such as melanin and bilirubin. In the presence of a thicker overlying tissue layer, such as severe subcutaneous edema or excess subcutaneous fat, it may be impossible for the NIR light beam to reach the target organ. In the newborn modest changes in weight have a small effect on abdominal measurements while changes in hemoglobin over the first weeks of life can change measurements by 30-50%. (Ferrari et al.,

nirswkshop1999.htm) Once NIRS monitors became commercially available a few animal and many clinical trials were conducted. The clinical investigations were for the most part small, brief observational prospective studies. Also NIRS was introduced into daily practice by others at that time, years before normative data and validation studies had been obtained.

There is great potential to use the NIRS technology in the neonatal intensive care unit (NICU) since it is a portable, continuous, non-invasive bedside monitoring technique. Following the development of small and skin friendly sensors and FDA approval of some NIRS monitors for use in neonates, both research and clinical use of NIRS in the NICU increased exponentially. The number of research projects over the last 5-10 years is large. However, the trials, while dealing with questions important to understanding physiology and clinical care in the NICU, are small and almost exclusively conducted at single centers. Often no more than 10-20 patients are being followed. Very large NIRS related studies enrolled 40-90 patients. Many of the observations reported are of brief sampling periods, sometimes being no more than spot samples.

This chapter is a limited overview for non-clinicians such as engineers and science students, or clinicians who want to learn about a medical application of NIRS. The recent introduction of the NIRS technology into neonatal medicine is used as an example of how a new device came into use into use in the clinical setting over the last decade. Main areas of clinical use and supporting studies will be mentioned. Limitations of NIRS technology and controversies as well as future directions will be addressed. With the abundance of available literature this chapter cannot claim to be a reference. This is an exciting and rapidly advancing field with new studies published even as this article was sent to press. This chapter will demonstrate how a new technology is adopted into medical care, in this case the NICU.

### **1.1 Materials**

Pub Med and Google have been queried regarding NIRS in NICUs, abdominal/splanchnic, cerebral and renal measurements, utility, and of NIRS use as prognosticator.

#### **1.2 Technology and measurements**

The principle of how NIRS works in humans was excellently summarized by Cohn:

Near-infrared spectroscopy has been used as a tool to determine the redox state of lightabsorbing molecules. This technology is based on the Beer-Lambert Law, which states that light transmission through a solution with a dissolved solute decreases exponentially as the concentration of the solute increases. In mammalian tissue, only three compounds change their spectra when oxygenated: cytochrome *aa3*, myoglobin, and hemoglobin. Because the absorption spectra of oxyhemoglobin and deoxyhemoglobin differ, their relative concentrations within tissue change with oxygenation, and the relative concentrations of the types of hemoglobin can be determined. Because NIRS measurements are taken without regard to systole or diastole, and because only 20% of blood volume is intra-arterial, spectroscopic measurements are primarily indicative of the venous oxyhemoglobin concentration. In the near infrared region (700 –1,000 nm), light transmits through skin, bone, and muscle without attenuation. (Cohn et al., 2003) There are several FDA approved NIRS monitors with somewhat different technology and algorithms available commercially (Wolf & Greisen, 2009) to measure the venous weighted regional oxygen saturation (rSO2) or tissue oxygenation index (TOI).

Due to the small size and the thin covering layers of tissue of both term and preterm neonates, r-SO2/TOI measurements at a depth of 2-3 cm can reach brain, kidney, gut/ splanchnic circulation, liver and muscle. The access to these critical organs promises valuable physiologic information through monitoring by NIRS. Measurements of several sites can be recorded simultaneously. (Hoffman et al., 2003; McNeill et al., 2010, 2011)

NIRS measurements are organ specific and regional (rSO2), reflecting perfusion and metabolism by non-invasive measurement in real-time. They are not temperature, pulsatility or flow dependent. Thus they may offer advantages over traditional measures of perfusion such as capillary refill, blood pressure, and urine output, lactate, venous and arterial O2 which tend to alert the clinician once the disease process is further progressed. R-SO2 measurements cannot stand alone. While they may often be the first sign of change, they need to be interpreted in the context of other measurements such as mean arterial blood pressure (MABP), pulse oximetry (O2sat), blood gases, additionally in the research setting with measurements of cerebral blood flow (CBF) and cerebral blood volume (CBV). Evaluation of the link between the venous weighted NIRS readings and peripheral pulse oximetry, a measure of arterial O2, gives insight into oxygen supply and demand. Using a simple equation, the fractional extraction of oxygen (FTOE = SaO2-rSO2/SaO2) oxygen consumption can be calculated and oxygen supply can be assessed. (Lemmers et al., 2006)
