Intracranial pressure (ICP) is affected in many neurological conditions. Clinical measurement of pressure on the brain currently requires placing a probe in the cerebrospinal fluid compartment, the brain tissue, or other intracranial space. This invasiveness limits the measurement to critically ill patients. Because ICP is also clinically important in conditions ranging from brain tumors and hydrocephalus to concussions, noninvasive determination of ICP would be desirable. Our model-based approach to continuous estimation and tracking of ICP uses routinely obtainable time-synchronized, noninvasive (or minimally invasive) measurements of peripheral arterial blood pressure and blood flow velocity in the middle cerebral artery (MCA), both at intra-heartbeat resolution. A physiological model of cerebrovascular dynamics provides mathematical constraints that relate the measured waveforms to ICP. Our algorithm produces patient-specific ICP estimates with no calibration or training. Using 35 hours of data from 37 patients with traumatic brain injury, we generated ICP estimates on 2665 nonoverlapping 60-beat data windows. Referenced against concurrently recorded invasive parenchymal ICP that varied over 100 millimeters of mercury (mmHg) across all records, our estimates achieved a mean error (bias) of 1.6 mmHg and SD of error (SDE) of 7.6 mmHg. For the 1673 data windows over 22 hours in which blood flow velocity recordings were available from both the left and the right MCA, averaging the resulting bilateral ICP estimates reduced the bias to 1.5 mmHg and SDE to 5.9 mmHg. This accuracy is already comparable to that of some invasive ICP measurement methods in current clinical use.
Further Reading TCD in ICP
Download fulltext: Non-invasive Methods of Estimating Intracranial Pressure
Transcranial Doppler Sonography
Because it is non-invasive, transcranial Doppler sonography (TCD) has been used for multiple applications, including detection of changes in cerebral blood ﬂow, vasospasm, circulatory arrest, and the evaluation of TBI [traumatic brain injury] . The non-invasive evaluation of ICP using TCD has also been studied. TCD generates a velocity–time waveform of cerebral blood ﬂow from which the peak systolic (PSV) and end-diastolic (EDV) ﬂow rates can be measured. The mean ﬂow velocity (MFV), resistance index (RI) or Pourcelot index—an indicator of resistance of an organ to perfusion, and the pulsatility index (PI) or Gosling index—a reﬂection of resistance encountered with the cardiac cycle are commonly reported derivations from the waveform display. (MFV = [(PSV + (EDV 9 x 2))/3], RI = (PSV - EDV)/PSV)], PI = (PSV - EDV)/MFV)
Klingelhofer et al. showed that ICP inﬂuenced TCD ﬂow patterns. With increasing ICP, there was a progressive reduction in the MFV and EDV with an increase in RI.
Using a mathematical model, Schmidt et al. were able to predict ICP trends based on arterial blood pressure and ﬂow velocity with a mean absolute difference of 4.0 ± 1.8 mmHg when compared to measured ICP. The above studies evaluated comatose or severely brain-injured patients with impaired cerebral autoregulation, where the PaCO2 was held constant at 30–35 mmHg, limiting the generalizability of the ﬁndings . Despite those limitations, in a series of 12 patients with severe head trauma, Goraj et al. were able to demonstrate a similar relationship between the RI and ICP trends, in patients with a PaCO2 range of 20–34 mmHg.
In a study of 81 adult and pediatric patients with a variety of intracranial pathologies, Bellner et al. concluded that there was a linear correlation between the PI and measured ICP, although the correlation decreased in sensitivity with higher ICP. Similar ﬁndings were demonstrated in a study of 125 patients with TBI by Moreno et al., where for each unit increase in ICP the PI increased by 0.03 unit.
In contrast, other groups failed to ﬁnd a relationship between PI and ICP. Behrens et al. used a lumbar infusion test to manipulate ICP in patients undergoing lumbar shunt placements. They found that the TCD blood ﬂow velocity was not a determinant of ICP. Similarly, Figaji et al. examined the relationship between the PI and ICP in 34 children with TBI who underwent 275 TCD studies. The absolute PI value was not a reliable correlate for ICP despite a weak tendency for higher PI values in elevated ICP. They recommended that PI need not be used to detect ICP unless the PI is very high.
Other studies have demonstrated equivocal results regarding the usefulness of TCD in adult and pediatric populations with progressive hydrocephalus . In addition, satisfactory TCD waveforms cannot be obtained in all patients.
Ultrasound of the ONS (optic nerve sheath)
Ultrasound of the ONS is a very well-studied modality for the non-invasive assessment of ICP. Ophthalmic ultrasound typically uses a frequency between 5 and 10.5 MHz to evaluate the eye and orbit [79, 80].
Hansen and Helmke  used ultrasound in a cadaver study to demonstrate that in the area just behind the eye-ball, elevated pressure can increase the sheath diameter by more than 50%. In another study, the same authors used intrathecal infusion tests to prove that the optic nerve sheath diameter (ONSD) varies with alteration of lumbar CSF pressure . A similar study was done by Tamburrelli et al., who showed that the ONS begins to expand when the diastolic ICP is increased to greater than 13–14 mmHg. Beyond that point, a linear correlation is seen between the enlargement of the ON sheath and simultaneous increases in ICP . These changes in the ON sheath occur before changes in the nerve are visible on fundoscopic examination . Using 4.5 mm as the cutoff for normal, Tamburrelli et al.  found a sensitivity of ultrasound to identify an ICP greater than 15 mmHg of 88% and a speciﬁcity of 90%.
Several studies have directly correlated ONSD measurements on ultrasound with ICP measured invasively.
The cut-off value for normal ONSD, measured 3 mm posterior to the globe, ranges from 5.2 to 5.9 mm. The sensitivity is 74–95% and the speciﬁcity is 74–100% to identify ICP >20 mmHg, as shown in Table 3 [85–90].
Ultrasound of the ONS has also been compared to ﬁndings of increased ICP on CT, such as changes in ventricle size, basilar cistern size, sulci size, degree of transfalcine herniation, and gray/white matter differentiation . Grisgin  showed that the mean ONSD in Neurocrit Care patients with signs of brain edema on CT is larger than that of controls. Other authors have shown that with a cut-off of normal ONSD ranging from 5.0 to 5.9 mm, the ONSD predicts ﬁndings of increased ICP on CT with a sensitivity of 74–100% and a speciﬁcity of 63–95%, as shown in Table 4 [92–95].
Studies of ultrasound to measure the ONSD have also been performed in children. Malayeri et al.  showed that there is a signiﬁcant difference between the ONSD in children with evidence of increased ICP on CT or transcranial ultrasound and the ONSD in controls. Two studies showed that the upper limit of normal is 4.0 mm in infants aged less than 1 year, and 4.5 mm in older children [97, 98]. Helmke and Hansen  agree that the ONSD should be considered enlarged when it is over 5.0 mm in children aged over 4 years. However, another study showed higher baseline values in asymptomatic patients with a history of hydrocephalus . Taking the upper limit of normal to be 4.0 mm in children aged less than 1 year and 4.5 mm over 1 year, Le et al.  found a sensitivity of 83% and speciﬁcity of 38% for predicting increased ICP as seen on CT or lumbar puncture opening pressure. Beare et al.  used a cut-off of 4.2 mm in children, and found a sensitivity of 100% and speciﬁcity of 86% for predicting increased ICP seen on CT.
Several studies have examined variables affecting the accuracy of ultrasound in measuring ONSD. Romagnuolo  showed that the ONSD does not change with patient position. Inter-observer variation is quite low , and the measurements are highly reproducible , even for novice operators taught in a single training session .
However, ultrasound is subject to artifacts , and measuring the ONSD in off-axis will result in an erroneous value . In order for ultrasound to be used reliably, a standardized technique must be used .
Color Doppler ultrasonography has also been used as a way to measure ICP non-invasively. The technique is similar to standard ultrasound, but the pressure on the globe must be minimized to prevent a decrease in blood ﬂow velocity . The ophthalmic artery and the central retinal artery and vein can reliably be detected [110, 111]. Doppler spectral analysis is accurate and reproducible , and blood velocity in the major retinal vessels varies minimally . Querfurth et al. showed that the arterial resistance rises with increasing ICP in the mild-moderate range; however, the ﬂow seems to normalize with more severe elevations of ICP. The authors speculate this may be due to local autoregulatory vascular changes and/or diversion of cerebral blood ﬂow into the ophthalmic circulation .
Miller et al.  showed that the central retinal artery systolic blood ﬂow velocity is signiﬁcantly reduced in children with increased ICP compared to controls. A study which combined ophthalmodynamometry to evaluate venous ﬂow and color Doppler to measure ﬂow in the ophthalmic and central retinal arteries showed that combining both parameters resulted in a better correlation with absolute ICP (r = 0.95) than either parameter alone .