RESUMO
Objective: This research aims to evaluate the possible association between pulsatile near infrared spectroscopic waveform features and induced changes in intracranial pressure in healthy volunteers. Methods: An optical intracranial pressure sensor was attached to the forehead of 16 healthy volunteers. Pulsatile near infrared spectroscopic signals were acquired from the forehead during body position changes and Valsalva manoeuvers. Features were extracted from the pulsatile signals and analyses were carried out to investigate the presence of statistical differences in the features when intracranial pressure changes were induced. Classification models were developed utilizing the features extracted from the pulsatile near-infrared spectroscopic signals to classify between different body positions and Valsalva manoeuvre. Results: The presence of significant differences in the majority of the analyzed features (p < 0.05) indicates the technique's ability to distinguish between variations in intracranial pressure. Furthermore, the disparities observed in the optical signal features captured by the proximal and distal photodetectors support the hypothesis that alterations in back-scattered light directly correspond to brain-related changes. Further research is required to subtract distal and proximal signals and construct predictive models employing a gold standard measurement for non-invasive, continuous monitoring of intracranial pressure. Conclusion: The study investigated the use of pulsatile near infrared spectroscopic signals to detect changes in intracranial pressure in healthy volunteers. The results revealed significant differences in the features extracted from these signals, demonstrating a correlation with ICP changes induced by positional changes and Valsalva manoeuvre. Classification models were capable of identifying changes in ICP using features from optical signals from the brain, with a sensitivity ranging from 63.07% to 80% and specificity ranging from 60.23% to 70% respectively. These findings underscored the potential of these features to effectively identify alterations in ICP. Significance: The study's results demonstrate the feasibility of using features extracted from optical signals from the brain to detect changes in ICP induced by positional changes and Valsalva manoeuvre in healthy volunteers. This represents a first step towards the non-invasive monitoring of intracranial pressure.
RESUMO
Photoplethysmography (PPG) signals from the forehead can be used in pulse oximetry as they are less affected by vasoconstriction compared to fingers. However, the increase in venous blood caused by the positioning of the patient can deteriorate the signals and cause erroneous estimations of the arterial oxygen saturation. To date, there is no method to measure this venous presence under the PPG sensor. This study investigates the feasibility of using PPG signals from the forehead in an effort to estimate relative changes in haemoglobin concentrations that could reveal these posture-induced changes. Two identical reflectance PPG sensors were placed on two different positions on the forehead (above the eyebrow and on top of a large vein) in 16 healthy volunteers during a head-down tilt protocol. Relative changes in oxygenated ( Δ HbO 2 ), reduced ( Δ HHb) and total ( Δ tHb) haemoglobin were estimated from the PPG signals and the trends were compared with reference Near Infrared Spectroscopy (NIRS) measurements. Also, the signals from the two PPG sensors were analysed in order to reveal any difference due to the positioning of the sensor. Δ HbO 2 , Δ HHb and Δ tHb estimated from the forehead PPGs trended well with the same parameters from the reference NIRS. However, placing the sensor over a large vasculature reduces trending against NIRS, introduces biases as well as increases the variability of the changes in Δ HHb. Forehead PPG signals can be used to measure perfusion changes to reveal venous pooling induced by the positioning of the subject. Placing the sensor above the eyebrow and away from large vasculature avoids biases and large variability in the measurements.
