In a preferred embodiment, the spectrophotometric device is configured to measure the attenuation at more than two light source to detector distances in order to improve the accuracy of the multi-distance method. In a preferred embodiment, the spectrophotometric device is configured to measure the attenuation at three light source to detector distances.

The distance between the light source and the light detector may be optimized based on, for example, the sensitivity of the detectors and the optical properties of the subject's tissue. In the case of a spectrophotometric device for measuring the blood oxygen saturation in the abdomen of a new-born infant, the shortest light source to detector distance is preferably at least 0.8 cm, more preferably at least 0.9 cm, and most preferably at least 1.0 cm. Preferably, the shortest distance between the light source and the detectors is in the range of 0.8 to 2 cm, more preferably at least 0.9 to 1.5 cm, and most preferably 0.95 to 1.2 cm. The longest light source to detector distance is preferably in the range of 2 to 10 cm, preferably 2.2 to 6 cm, most preferably 2.5 to 5 cm.

The at least one light source and the at least one light detector may be configured such that they can be brought into direct contact with the subject's skin in order to avoid any interference with ambient light.

The one or more light sources and the one or more light detectors are preferably positioned in a single housing. The housing is preferably configured to be attached to or positioned on the subject's skin. More specifically, the housing is preferably configured to be attached to or positioned on the subject's abdomen.

The housing may be made from any suitable material, preferably from a plastic material. The housing has at least one surface facing the subject's skin, and the one or more light sources and the one or more light detectors are positioned on this surface. Preferably, the housing has a single surface facing the subject's skin and both the light sources and the light detectors are positioned on this surface.

The surface facing the subject's skin may be planar or it may be shaped to correspond to the surface of the body part, on which the spectrophotometric device is placed. For example, the surface of the housing may be concavely curved to correspond to the convex curvature of a subject's abdomen.

In a preferred embodiment, the housing is provided with an adhesive patch for attaching the housing to the subject's clothing. The adhesive patch may preferably be provided on a surface of the housing facing away from the subject's skin, preferably on a surface opposite the abovementioned surface on which the light sources and light detectors are provided. For example, the adhesive patch may be used to attach the housing to a diaper worn by the subject. Thereby, the spectrophotometric device can be fixed underneath the subject's clothing, such as a diaper. The adhesive patch is preferably a Velcro patch or hook-and-loop fastener.

The housing may be made from a rigid material to that the distance between the light source and the detector remains constant. This may be particularly preferable, if the spectrophotometric device uses the multi-distance method for calculating the blood oxygen saturation. Alternatively, the housing may be made from a flexible material. This has the advantage that the shape of the housing can be adapted to the individual subject and the shape of the individual body part. This allows a better fit between the spectrophotometric device and the subject's skin and may help to eliminate measurement artifacts arising from ambient light hitting the light detector.

According to the first aspect, the spectrophotometric device further comprises a motion sensor. The motion sensor generates a motion signal indicative of motion of the light source and/or the light detector. Therefore, the motion sensor is preferably positioned in close proximity to the light source and/or light detector. The motion sensor may be provided in the above-described housing together with the one or more light sources and the one or more light detectors. In this case, the motion signal is preferably indicative of movement of the entire housing. However, it is also possible that the motion signal is indicative specifically for motion of one of the light sources and/or one of the light detectors.

The motion sensor may comprise, for example, a linear accelerometer configured to measure linear displacement or a gyroscope configured to measure rotational displacement. A combined linear accelerometer and gyroscope may also be used to measure both linear displacement and rotational displacement. Preferably, the motion sensor is piezoelectric device or a microelectromechanical system (MEMS).

Preferably, a linear accelerometer is used. A suitable linear accelerometer may measure displacement along three axes, two axes or a single axis. Most preferably, a linear 3-axis accelerometer is used.

The motion sensor preferably has a sampling rate of 1 kHz to 5.3 kHz. The motion sensor preferably measures acceleration in the range of ±16 g, more preferably ±8 g, most preferably ±4 g, more preferably ±2 g.

The electronic data processing unit configured to calculate the blood oxygen saturation is configured to receive data from both the motion sensor and the one or more light sources and one or more light detectors. The electronic data processing unit may be physically connected to the motion sensor, the light sources and light detectors, e.g. by an electrical wire connection. However, the electronic data processing unit may also be positioned at a remote location and be connected wirelessly to the motion sensor, the light sources and light detectors. The electronic data processing unit may also be a central processing unit that is part of a multitude of spectrophotometric devices. In this case, the electronic data processing unit receives signals of a multitude of motion sensors, light sources and light detectors to calculate blood oxygen saturation for a multitude of subjects.

The electronic data processing unit may be implemented, for example, on a personal computer comprising a central processing unit, a random access memory, a data storage device and a data connection for receiving the measurement data from the motion sensor and the light detectors.

The electronic data processing unit carries out an algorithm for calculating the blood oxygen saturation. In this context, blood oxygen saturation StO₂is defined as the ratio of the concentration of oxyhemoglobin HbO₂to the total concentration of oxyhemoglobin HbO₂to deoxyhemoglobin Hb according to the following equation:

StO₂=c(HbO₂)/(c(HbO₂)+c(Hb))

where c(HbO₂) is the concentration of oxyhemoglobin and c(Hb) is the concentration of deoxyhemoglobin.

The algorithm involves determining the light attenuation A_λ of the light signal emitted by the one or more light sources and detected by the one or more light detectors as a function of the wavelength λ. From the light attenuation A_λ, the algorithm then calculates the concentrations of oxyhemoglobin c(HbO₂) and the concentration of deoxyhemoglobin c(Hb) based on the Lambert-Beer Law. Specifically, the algorithm uses empirically determined extinction coefficients of oxyhemoglobin and deoxyhemoglobin and empirically determined calibration data accounting for light scattering in the subject's tissue. The extinction coefficients of oxyhemoglobin and deoxyhemoglobin can be determined from absorption spectra of oxyhemoglobin and deoxyhemoglobin in isolation. Preferably, the algorithm only determines relative concentrations of oxyhemoglobin and deoxyhemoglobin. As the blood oxygen saturation is defined as the ratio of the concentrations of oxyhemoglobin and deoxyhemoglobin, it is not necessary to determine the absolute concentrations.

In one embodiment, the algorithm involves calculating the concentrations of oxyhemoglobin c(HbO₂) and the concentration of deoxyhemoglobin c(Hb) using the differential wavelength method. This involves measuring the light attenuation at least at three distinct wavelengths and calculating the difference in attenuation between the wavelengths. The differential wavelength method is described, for example, in EP 1 259 791 B1.

In another embodiment, the algorithm involves calculating the concentrations of oxyhemoglobin c(HbO₂) and the concentration of deoxyhemoglobin c(Hb) using the multi-distance method. This involves measuring the light attenuation at two or more different light source to detector distances, calculating the slope of the attenuation of the light signal versus the light source to detector distance as a function of the wavelength and calculating the blood oxygen saturation within the subject's tissue on the basis of said slope of the attenuation of the light signal. The multi-distance method is described, for example, in Tachtsidis, Ilias et al. “A Hybrid Multi-Distance Phase and Broadband Spatially Resolved Spectrometer and Algorithm for Resolving Absolute Concentrations of Chromophores in the Near-Infrared Light Spectrum.” Advances in Experimental Medicine and Biology 662 (2010): 169-175.

By measuring the attenuation as a function of the light source to detector distance, it is possible to calculate the relative absorption μ_α,λ using the multi-distance approach. This removes the necessity of performing calibration by determining the blood oxygen saturation of a given reference tissue assuming that oxygen saturation of said reference tissue is the weighted sum of the oxygen saturation of a subject's venous and arterial blood. This method only needs to account for light scattering of the tissue. This removes a source of systematic error, since it no longer requires any assumptions on the relative contributions of venous and arterial blood.

The electronic data processing unit may further be configured to determine the subject's pulse and/or breathing frequency using the pulsatile part of the measured light attenuation. In this case, it is necessary that the electronic data processing unit continuously samples the light signal detected by the light detector over a given period of time.

In a preferred embodiment, the algorithm takes into account absorption from other chromophores in addition to oxyhemoglobin and deoxyhemoglobin. The algorithm preferably takes into account empirically determined data that account for wavelength dependent light absorption by light absorbers other than hemoglobin. Preferably, the algorithm takes into account light absorption due to other light absorbing species in a subject's abdomen. Most preferably, this includes absorption by stool, transitional stool of infants, meconium, and/or billiverdin.

In one particular embodiment, the algorithm takes into account the wavelength dependent extinction coefficients of these other light absorbers. The extinction coefficients can be determined by measuring the absorption spectra of, for example, isolated samples of stool, transitional stool, meconium, and/or biliverdin. In a preferred example, the extinction coefficients are determined by measuring the absorption spectra of isolated samples of meconium.

Meconium is the earliest stool of a mammalian infant. Meconium is composed of materials ingested during the time the infant spends in the uterus: intestinal epithelial cells, lanugo, mucus, amniotic fluid, bile, and water. It has been found that averaged absorption spectra of meconium samples taken from number of different subjects can be used as a source of extinction data for the above calculation. In one embodiment, the data accounting for attenuation of the light signal due to light absorbers therefore include the wavelength-dependent extinction coefficients of meconium samples taken from new-born infants.

Transitional stool is produced by a new-born infant during its first days of life. Transitional stool differs from meconium in its composition and comprises high amounts of biliverdin. Therefore, the data accounting for attenuation of the light signal due to light absorbers preferably include the wavelength-dependent extinction coefficients of transitional stool samples taken from new-born infants, preferably during the first two weeks after birth, more preferably during the first week after birth, most preferably during the first five days after birth.

In another preferred embodiment, the data accounting for attenuation of the light signal due to light absorbers therefore include the wavelength-dependent extinction coefficients of biliverdin.

The electronic data processing unit is preferably configured to continuously sample the measurement data of the one or more light detectors and to calculate the blood oxygen saturation. The sampling rate preferably is in the range of 0.1 kHz to 10 kHz, more preferably 0.5 kHz to 3 kHz, most preferably 0.6 kHz to 2 kHz. The electronic data processing unit may also perform some form of downsampling, for example by calculating an average light signal from a group of consecutive, discrete measurements provided by the one or more light detectors. Preferably, the electronic data processing unit calculates a moving average of the measurement data of the one or more light detectors and calculates the light attenuation and the blood oxygen saturation based on this moving average.

The spectrophotometric device may comprise an electronic storage device connected to the electronic data processing unit, which stores data identifying the spectrophotometric device and calibration data specific for the spectrophotometric device. During operation, the electronic data processing unit can use this information to calculate the blood oxygen saturation. Preferably, the electronic storage device is an integral part of the electronic data processing unit. However, this is not essential. The electronic storage device does not necessarily have to be positioned at the same location or in proximity of the electronic data processing unit. The electronic storage device may also be provided in the above-mentioned housing containing the light sources and detectors, or it may be positioned in a connector connecting the housing to the electronic data processing unit. The electronic storage device may further be positioned at a remote location and may be connected to the electronic data processing unit, for example, through a wireless location.

According to the first aspect of the invention, the electronic data processing unit is configured to discard the light signal measured by the at least one detector, if the motion signal detected by the motion sensor exceeds a predetermined threshold value. This improves the accuracy of the measurement and reduces false alarms, as measurement artifacts caused by movement of the spectrophotometric device and/or the subject can be identified and discarded.

The electronic data processing unit is preferably configured to continuously sample the motion signal provided by the motion sensor. This preferably occurs at the same sampling rate as the sampling of the measurement of the one or more light detectors. The electronic data processing unit may perform downsampling of the measurement data received from the motion sensor as described for the light signal received by the light detector above, e.g. by calculating a moving average of the motion signal.

The electronic data processing unit may be configured to continuously sample the measurement data of the one or more light detectors to calculate the blood oxygen saturation and to continuously sample the motion signal provided by the motion sensor. If the motion signal exceeds the predetermined threshold in this embodiment, calculation of the blood oxygen saturation is interrupted.