Investigation of oxygen saturation derived from cardiac pulsations measured on the adult head using NIR spectroscopy.

Cardiac related pulsatile signals can be detected in different parts of the human body, including the finger, ear lobe and forehead [1] by using near infrared (NIR) monitoring. These pulsatile signals are due to attenuation of light by the increase of arterial blood volume during systole in the cardiac cycle. Pulse oximetry exploits these pulsatile signals to calculate oxygen saturation (SpO2) [2]. There are two types of pulse oximetry: (1) the transmission type where the light source and detector are facing each other across the measurement site (e.g. ear lobe or finger), and (2) the reflectance type where both the light source and detector are in the same plane (e.g. forehead or forearm) [1] with the source detector spacing typically less than 1 cm. With more sensitive optical instruments, it has been shown that these pulsatile signals can be measured on the forehead at a greater spacing using either a CCD spectrometer [3] or a phase resolved system [4]. Both of these two methods can be considered as operating in reflectance mode with large source detector spacings of 3 or 3.5 cm. These pulsatile signals were also thought to be mainly caused by the change in arterial blood volume. However, at source detector spacing larger than ~2 cm, NIR light can penetrate through the skull into the brain and the measured pulsatile signals are likely partly to include components caused by brain movement [5] as well as arterial blood pulsations in the scalp and the brain. The objective of this paper is to compare three algorithms used to calculate oxygen saturation from the head pulsatile signals, (signified by SpO2 to distinguish it from SpO2 measured at other sites) with large source detector spacing. Two of the algorithms implicitly allow the possibility of venous blood contributing to SpO2. We will show the SpO2 calculated by three algorithms in 8 adult subjects during normoxia and hypoxia. Examples of phase differences between the oxy and deoxy-haemoglobin (∆HbO2 and