Real-Time Monitoring of Neurocritical Patients with Diffuse Optical Spectroscopies

Monitoring brain physiology noninvasive in real-time during neurocritical patient treatment may reduce mobility and mortality rates, which is particularly critical in remote areas and open fields. With the rise in portable and relatively cheap instrumentation, diffuse optics can be used to measure blood flow and oxygenation simultaneously, with a high temporal resolution and without critical care interference. The real-time feature has the potential to guide treatment strategies for brain injuries in the neuro ICU basis of patient-specific cerebral physiology, rather than universally applied standard protocols or surrogate metrics.

Before moving the system into the ICU, connect all the fibers to the relevant lasers and detectors and make sure they are properly attached to the optical probe. Then, obtain consent from either the patient or a legal representative and bring the cart to the patient's room. Upon arrival, turn off the DCS laser.

And in the File menu, click Load Settings File to load the appropriate probe settings file. Place the probe on the curved side of the calibration phantom, ensuring good contact with the surface, and click Optimize All Detectors to optimize the PMT bias voltage. In the Calibrate menu, click Calculate Waveform Calibration Values For Optical Props and Multiple Distances to run the calibration for multiple source detector separations.

In the Text-Mon menu, open the user-defined option to check that the measured optical properties match the pre-specified values and that the fitting of R-squared is close to one. Then, probe the phantom again to measure the optical properties of the phantom marked as check to ensure that the calibration was adequate. The measured optical properties should be within 10%of the values specified in the phantoms.

To prepare the patient for the analysis, first use sanitizing wipes to clean both the probe and the patient's forehead. Place double-sided tape on the probe, taking care that the tape is not in direct contact with the optical fiber tips, and place laser safety goggles onto the patient. Place the probes over the region of interest and wrap the elastic straps around the patient's head.

Then, turn on the DCS laser. Before beginning the data acquisition, enter the DCS source detector separations in the Settings tab of the graphic user interface. To begin the acquisition, press Start in the system software.

Click Optimize All Detectors to optimize the PMT bias voltage for the patient, and check the optical properties and R-squared of the DOS fitting as demonstrated. Turn on the DCS detectors and check that each DCS detector is measuring an adequate light intensity. Typically, more than 10 kilohertz are required.

If the measured intensity exceeds 800 kilohertz, use a neutral density filter to reduce the photon counts to avoid damaging the detectors. Check the auto-correlation curves in the Correlation Curves tab of the graphic user interface to ensure a good skin coupling, and reposition the optical probe, if necessary. When a good contact between the probe and the skin is achieved, click Stop to stop the data collection and set the experiment and patient identifiers in the Folder text box and the region of interest name in the File name text box.

Then, press Start to start the data acquisition. After starting the measurement, enter the relevant patient information in the Experiment Info tab and click Mark to ensure that any relevant event that occurs during the monitoring period is marked. After each mark, enter the event description in the Experiment Info tab.

At the end of the assessment, press Stop in the graphic user interface to stop the data collection. Turn off the DCS detector switch and turn the DCS laser key to the off position. Click all detectors off to turn off the PMTs and remove the probe from the patient's head and the double-sided tape from the probe.

Then, clean the probe and the system and accessories with sanitizing wipes. Normalized auto-correlation curves obtained with the DCS module should be approximately 1.5 at the zero delay time extrapolation and should decay to one at longer delay times. In this example of a bad auto-correlation curve, it is not possible to distinguish the curves for the different source detector separations.

While in this example, the curve's tail did not match the model used. Here, the time series from a monitoring session for a sedated stroke patient during which the patient's bronchial and oral secretions were suctioned can be observed. The patient motion induced by the intervention clearly disturbed the optical signal, leading to unphysiological spikes in the optical parameters.

Soon after the intervention, the hemodynamic parameters returned to approximately the same values as before the intervention, as expected for a stable patient. During this monitoring session, 13 days after hospitalization for ischemic stroke, the cerebral blood flow and cerebral metabolic rate of oxygen in the ipsilesional forehead were considerably lower than their contralesional parameters in the symmetric region, while a high oxygen extraction fraction was observed in both hemispheres. Careful analysis of this representative laterality index data of a different patient revealed one significant period of impairment between the first and the third days after hospitalization and a second period starting after the third day of hospitalization.

Using continuous arterial blood pressure monitors, our protocol can be slightly modified to also assess cerebral regulation in neurocritical patients, providing clinical relevant data. The potential to assess each patient's neurophysiology at the bedside can help us to better understand the progress of different injuries and impact of different courses of care.