Noninvasive Monitoring of Intracranial Pressure

Authors: Djordje Popovic1,3,*, Michael Khoo1 and Stefan Lee 2

(1Department of Biomedical Engineering, Viterbi School of Engineering, University of Southern California, Los Angeles, CA, 2 Department of Neurosurgery, Keck School of Medicine, University of Southern California, Los Angeles, CA, 3 Advanced Brain Monitoring Inc., Carlsbad, CA)

Recent Patents on Biomedical Engineering 2009, 2, 165-179
Received: April 29, 2009; Accepted: May 6, 2009; Revised: May 9, 2009

Increased intracranial pressure (ICP) is one of the major causes of secondary brain ischemia that accompanies a variety of pathological conditions, most notably, traumatic brain injury (TBI), stroke, and intracranial hemorrhages. However, aside from a few Level I trauma centers, ICP monitoring is rarely a part of the clinical management of patients with these conditions because of the invasiveness of the standard monitoring methods (which require insertion of a catheter into the cranium), additional risks they present for patients, high costs associated with the procedure, and the limited access to trained personnel, i.e., a neurosurgeon. Alternative methods have therefore been sought with which ICP can be measured noninvasively. This article reviews nearly 30 such methods patented over the past 25 years, which included ultrasound “time-of-flight” techniques, transcranial Doppler, methods based on acoustic properties of the cranial bones, EEG, MRI, tympanic membrane displacement, oto-acoustic emission, ophthalmodynamometry, and ultrasound measurements of optic nerve sheath diameter. At present, none of the methods is sufficiently accurate to allow for routine clinical use although several hold promise. Future developments should integrate further refinements of the existing methods, combined use of multiple sensors and/or technologies, and large clinical validation studies on relevant populations.  


Intracranial pressure (ICP) is the pressure inside the skull, or more specifically, the pressure of cerebrospinal fluid (CSF) that fills in the four brain ventricles and the subarachnoid space around the brain Fig. (1A). Its value, conventionally referenced to the atmospheric pressure and expressed in millimeters of mercury, is maintained in a narrow range, between 7 and 15 mmHg for adults, 3 - 6 mmHg in children, and 1.5–6mmHg in term infants . Because the skull is a rigid container and its content incompressible, ICP is essentially determined by volume equilibrium among the brain (1,100-1,300cm 3), CSF (130-150cm3 )and blood contained in intracranial vessels (60-80cm3). Any increase in volume of one constituent must be accompanied with a reduction in volume of other constituents if ICP is to remain constant (“the Monroe-Kelly doctrine”). Given that the brain volume is fixed, the two most important determinants of ICP under physiological conditions are the cerebral blood flow, which is normally tightly regulated and remains nearly constant for a wide range of the mean arterial blood pressures , and the balance between the production of CSF in the choroid plexus of the brain ventricles and its absorption into the dural venous sinuses. 

Increased ICP, known also as intracranial hypertension (IH),is therefore synonymous with volume expansion in the intracranial compartment. The volume expansion can be caused by a variety of reasons that include intracranial hemorrhages, brain tumors, localized or generalized brain swellling, and obstruction to CSF flow or absorption Table 1. The exact relationship between intracranial volume changes and ICP is described with a sigmoidal curve Fig. (1B). Volume expansion of up to 30cm3 usually results in insignificant changes of ICP because it can be compensated by extrusion of CSF from the intracranial cavity into the reservoir of the spinal theca, and, to a lesser extent, extrusion of venous blood from the cranium (region I in Fig (1B). When the compensatory mechanisms have been exhausted, ICP rises rapidly with further increases in volume (region II in Fig. (1B)) until it reaches the level comparable with the pressure inside of cerebral arterioles (which depends on mean arterial blood pressure and cerebrovascular resistance but normally measures between 50 and 60 mmHg). At this point the rise of ICP is halted as cerebral arterioles begin to collapse and the blood flow completely ceases (region III - the terminal plateau of the curve in Fig. (1B).  The major consequences of intracranial hypertension are derangement of cerebrovascular reactivity and impediment of cerebral blood flow with consequent ischemia of the affected brain tissue. Since the pressure in large intracranial veins and venous sinuses is only slightly (2-3 mmHg) higher than ICP, cerebral blood flow (CBF) is effectively determined by three factors: mean arterial blood pressure (MAP), ICP and cerebrovascular resistance (CVR), according to the following Eq. (1): 

  Eq. (1)

 Initially, the negative effects of elevated ICP on cerebral blood flow are first countered by a compensatory decrease in cerebrovascular resistance through reflex vasodilatation of

    Table 1. Causes of Intracranial Hypertension (Grouped by Mechanism) 


  Fig. (1). (A). Anatomy of the intracranial compartment; (B). Intracranial pressure-volume curve (based on unpublished data from piglets). 

cerebral arterioles, and, if this does not suffice, with an increase of mean arterial blood pressure. Unfortunately, both compensatory mechanisms result in an effective expansion of cerebral blood volume and, consequently, further increase the ICP. Eventually, the reactivity of cerebral arterioles is lost (arterioles have been maximally dilated), cerebral blood flow begins to decrease, and a vicious cycle establishes in which the increased ICP impedes cerebral blood flow, leading to brain tissue ischemia with consequent swelling, which through a volume expansion further increases ICP. This amplification mechanism is one of the main factors responsible for the development of a secondary, generalized ischemic brain injury within hours or days from the onset of the primary condition that caused intracranial hypertension.  

Another dire consequence of high ICP is a shift of brain masses, which typically occurs after a large focal mass lesion (e.g. tumor, hemorrhage, or edema) that pushes away the nearby brain structures across the midline, or towards the tentorial hiatus and/or foramen magnum Fig. (1A). These shifts may result in additional vascular and neuronal injuries, disruption of the flow of CSF (which further increases ICP), and a fatal compression of the brainstem.


The standard technique for direct ICP monitoring is an invasive procedure that involves inserting a catheter into the intracranial compartment, usually through a small burr hole, and connecting it to a standard pressure transducer. Lateral ventricle is the preferred location as it provides the global ICP, uninfluenced by eventual pressure gradients in the brain parenchyma, and allows for therapeutic drainage of CFS and administration of drugs . Placement of the intraventricular catheter requires its penetration through the brain cortex and underlying white matter, and may be difficult in the presence of brain swelling or intracranial mass lesions that may affect the shape and/or location of the lateral ventricles. Infections complicate the use of intraventricular catheters in 1 to 11% of cases, with the risk increasing with the duration of continuous use or with repeated uses . Newer catheters may be impregnated with antibiotics which lowers the infection rate . Catheters in locations other than the lateral ventricles (e.g. epidural or subarachnoid catheters) are now rarely used because of the much lower accuracy of such measurements . Spinal tap with a needle inserted between the lumbar vertebrae and connected to a pressure transducer via a catheter is attractive because it has low complication rates and can be performed as ambulatory procedure. The CSF pressure at the lumbar level is however not a reliable estimate of the pressure inside the cranium. Additionally, the procedure is contraindicated if very high ICP is suspected because it may provoke a herniation with consequent brain stem compression and death.  Alternatively, microtransducer-tipped ICP monitors can be inserted in the brain parenchyma or subdural space through a skull bolt, small burr hole, or during a neurosurgical procedure . The pressure can be transduced with optic fibers (e.g. Camino ICP monitor, Integra Neuroscience, Plainsboro, NJ ), strain gauge (e.g. Codman Microsensor, Raynham, MA , or Neurovent-P ICP monitor, Raumedic AG, Munchberg, Germany ) or pneumatic technologies (e.g. Spiegelberg ICP monitor, Spiegelberg GmbH, Hamburg, Germany ). The insertion is technically simpler because of the more superficial placement of the catheter, and the risk of infections or other complications is minimal . However, measured pressure may not be representative of the true ‘global’ ICP because pressure gradients that may exist in the brain parenchyma, especially after a focal, unilateral mass lesion. The output of fiberoptic or strain-gauge tipped pressure transducers steadily drifts, which usually necessitates their replacement because the transducers cannot be recalibrated in vivo. Recent advances in manufacturing and testing have however significantly reduced the drift, enabling that the transducers be used reliably for more than one week . Pneumatic transducers such as Spielberg monitor can also be zeroed in vivo, which eliminates the drift, but their narrow bandwidth allows only mean ICP to be recorded, whereas the analysis of ICP waveforms is not possible. Implantable ICP monitors have also emerged with a catheter-tip transducer connected to a telemetry unit implanted into a skin pouch outside of the cranium. Telemetric monitors are typically integrated with ventriculo-peritoneal shunts, and used for long-term monitoring of ICP in patients with hydrocephalus . In summary, the invasive methods for monitoring ICP share several common drawbacks Table  2. Pressure sensors are subject to a steady zero-drift which in some cases is significant and requires in-vivo recalibration (where possible) or replacement of the sensor. Insertion of a catheter or transducer carries a risk of brain or spinal cord damage and infection which increases with repeated insertions or frequent manipulations of the catheter. Finally, the insertion can be done only by a highly trained individual (neurosurgeon, or anesthesiologist in case of the spinal tap). Invasive ICP monitoring is therefore in practice limited to neuro-intensive care units and specialized hospitals, and cannot be routinely used in general hospitals, emergency rooms, ambulatory settings or in the field.


Two most frequent causes of intracranial hypertension, traumatic brain injury (TBI) and stroke, are worldwide epidemics with an estimated joint annual incidence of between 400 - 600 cases per 100,000 inhabitants in the developed countries and mortality of 30-50%.

   Table 2. Methods for Invasive Monitoring of ICP

Furthermore, about one third of TBI and nearly half of stroke survivors suffer from permanent functional disabilities whose degree depends on the extent and severity of damages to the brain tissue: in the US only this translates into more than 6 million people . Intracranial hypertension is a primary cause of secondary ischemic brain injury after either condition, and its degree and duration have been shown to correlate with the survival rates , and severity of permanent functional disabilities . Prevention and control of intracranial hypertension are therefore the fundamental goals in the management of patients with TBI or stroke, and ICP monitoring might serve as a basis for making therapeutic decisions as well as provide an objective measure of success of the applied treatment.  

Continuous monitoring of ICP is recommended by consensus guidelines of the American Association of Neurological Surgeons (AANS) and Brain Trauma Foundation (BTF) in patients with severe TBI as a means to guide therapeutic interventions and assess prognosis. Treatment guided by mean ICP or CPP was shown to have resulted in decreased mortality and shorter hospital stays , and improved functional outcome in survivors as compared to patients in whom therapy was guided only by standard clinical monitoring. Some studies however found no significant advantage of ICP- or CPP-driven treatment protocols . In practice, the use of ICP monitoring varies from one center to another, and considerable reserve exists among clinicians with respect to its utility. Large epidemiologic surveys from 2001-2002 revealed that ICP was monitored in only 58% of patients in the US and 37% in Europe in whom the monitoring was indicated, whereas in a Canadian study (where ICP monitoring was almost universal) only 20% of neurosurgeons relied on ICP when making therapeutic decisions. More recently, higher compliance with the AANS/BTF recommendations has been reported for Level 1 trauma centers in the United States (77% on average) , but this still means that 1 out of 5 patients in whom the monitoring is indicated does not get it. Such attitude may be caused by the lack of decisive Class 1 evidence of the utility of ICP monitoring, but at least part is motivated solely by the risks associated with the ICP monitoring. 

ICP is rarely monitored in patients with other conditions accompanied with intracranial  hypertension, including stroke. The main reason for such state of affairs is that the risks and cost of the procedure significantly outweigh uncertain benefits it provides in terms of the treatment guidance and improved clinical or functional outcome. Therefore, most clinicians including neurosurgeons prefer to rely on a combination of brain imaging (CT, MRI) and clinicalobservations that include the subjective evaluation of symptoms and signs suggestive of IH such as headaches, confusion, papilledema or worsening neurological deficits. Exceptions are limited to patients in whom an intracranial catheter is primarily a therapeutic tool, such as patients with hydrocephalus who have been implanted with a permanent catheter (shunt) that drains CSF into the abdominal cavity, or patients temporarily implanted with a drainage catheter after a major brain surgery. It should be emphasized here that the aforementioned clinical signs of intracranial hypertensions are  late, i.e. they manifest after the ICP has been substantially increased for considerable time (1 – 2 hours), and some degree of brain ischemia may have already developed. 


The use of (invasive) ICP monitoring in clinical practice is suboptimal because of the three main reasons: 

1. The insertion of a catheter or transducer is traditionally done only by a neurosurgeon in a specialized facility. Many patients with acute intracranial hypertension are however treated in intensive care units of general hospitals or other medical centers that do not necessarily have a neurosurgeon. Although emerging data suggests that physicians who are not neurosurgeons, or even physician assistants could place subdural or parenchymal transducers with low complication rates, such practice would still be considered problematic in most intensive care units. As a result, ICP monitoring is not performed in many patients in whom it is indicated.

2. The risks associated with the procedure frequently outweigh the value of added diagnostic information. 

3. Powerful neuro-imaging techniques (CT, MRI) are readily available that, in addition to providing other diagnostic information, allow for assessment of edema, bleeding, intracranial masses, or signs of increased ICP, and can therefore partially substitute for direct ICP monitoring. The signs of raised ICP on CT or MRI scan (attenuated visibility of sulci and gyri, changes in size of the lateral ventricles, poor distinction between the gray and white matter, and compression of the suprasellar and quadrigeminal cistern) are however only qualitative
indicators, and may be inaccurate, especially if ICP has been chronically increased or slowly increasing so that the brain structures have had time to adapt. An assessment of ICP based on repeated CT or MRI scans would therefore be inefficient, expensive, and potentially dangerous for the patient (because neither CT nor MRI are allow for bedside assessment, which would require frequent transportation of the patient from the IC unit to the imaging facility).

 A method of ICP monitoring that requires no surgery and poses no risks of infection or hemorrhage would certainly be welcomed by medical professionals, and at the very least could become the new ‘gold standard’ in neuro-intensive care units provided it is sufficiently accurate and easy to use. If the method is relatively inexpensive and does not require a specialist in order to be applied correctly, or can be automated to some degree, it could also find its way to various levels of the health care system with a potential to substantially modify the current concepts of management of patients with conditions accompanied with intracranial hypertension, including but not limited to TBI. Some examples include:

1. Triage at the Point of Contact

Noninvasive assessment of ICP could be applied during the first few hours after a brain injury or insult at the points of contact with limited human and technological resources (e.g. rural medical centers, community clinic) or in the field (e.g. wounded soldiers in the battlefield) and serve to identify patients that need further diagnostic examination, prompt treatment or transportation to a specialized facility. The ultimate result would be a more rational and cost-effective use of the available resources. 

2. In-Time and Evidence-Based Application of Therapy

At present, therapy for intracranial hypertension is frequently administered blindly, i.e. without an exact knowledge of ICP or insight into the functional state of cerebral vasculature. Universal application of noninvasive ICP monitoring in neuro-intensive care units would allow for timely initiation and modification(s) of treatment, which should, at least in theory, result in improved survival rates and functional outcomes. However, (and in conjunction to what has been said about triaging patients at points of contact) availability of noninvasive ICP monitors at medical centers that lack expertise on how to use this information most effectively may be dangerous because of serious side effects of ICP/CPP-driven therapeutic protocols . Such protocols are therefore likely to remain reserved for Level 1 trauma centers only, where the expertise is available. 

3. Long-Term Monitoring

Currently, ICP monitoring is usually discontinued after a few days because of the increased risk of infections (intraventricular catheters) or considerable drift of the transducer (subdural and intraparenchymal monitors). Noninvasive monitoring could be continued into the subsequent phases of recovery and serve to predict the speed and degree of recovery of the impaired functions, which may assist in planning rehabilitation. 

4. New Indications

Noninvasive ICP monitoring could become a part of the routine assessment of patients with other frequent conditions accompanied with intracranial hypertension, most notably in the acute management of stroke survivors or chronic management of patient with hydrocephalus, idiopathic intracranial hypertension, or slowly growing brain tumors. It could also provide a cost-effective method for clinical trials designed to assess new treatments for TBI and stroke.


Methods for noninvasive ICP monitoring estimate ICP from physiological or anatomical characteristics that are influenced by intracranial pressure and at the same time can be measured directly and non-invasively. The methods that have been proposed so far can be broadly divided in two clusters: those that derive ICP from anatomical or functional properties of the structures within the cranium (cranial bones, brain tissue, CSF, intracranial blood vessels), and those that infer ICP from an anatomic and/or functional property of extracranial organs that are anatomically connected or functionally linked with the intracranial compartment. Methods in each cluster can be further grouped on the basis of principles and technologies used for the measurement Table  3. All the methods will be described in the remainder of this section of the article, and analyzed with respect to the six characteristics important for the clinical utility and commercial success (ordered by clinical importance):

1. Accuracy 

How well does the method compare against the invasive ICP measurement? Does the method quantify ICP on a continuous scale (in mmHg) or it provides a few categories based on ranges (e.g. <20mmHg, 20–40mmHg, and >40mmHg)? The Association for Advancement of Medical Instrumentation (AAMI) has for example specified that ICP monitoring devices should have continuous output in the 0-100mmHg range with an accuracy of ±2mmHg in the 0-20mmHg range, and maximum error of 10% for ICP above 20mmHg, the specifications supported by the Brain Trauma Foundation guidelines .  

   Table 3. Methods for Noninvasive ICP Monitoring

2. Limitations / Contraindications

What are the physiological or pathological conditions in which the method will give erroneous readings, or in which it might be contraindicated? How frequent are these conditions in the general population, and how often might they be expected to coincide with intracranial hypertension?

3. Ease of Use 

How cumbersome and time consuming is it to apply the sensor(s) and transducers(s) utilized by the method? What level of skills and medical knowledge is needed in order to perform meaningful measurements?

4. Costs of Use

Comprised of the cost of technology, and costs of infrastructure required for the correct use of the technology. The costs will be evaluated on a somewhat arbitrary 3-item scale: low (measured in hundreds of dollars), moderate (in thousands) and high (in tens or hundreds of thousands). 

5. Portability and Field Deployability

Is the required equipment light enough to be easily transported by hand, and can it be battery powered for a reasonable time? Can it be applied outside the health care facilities and by personnel with limited medical education?

6. Continuous Monitoring

Does the apparatus allow for continuous monitoring, or repeated measurements are required? This question has two dimensions: practical, that pertains more to the ease of use, and diagnostic, that is related to the availability of continuously recorded ICP waveforms whose morphology carries information about intracranial compliance that is not available from point-measurements of ICP at regular time intervals. 

5.1. Methods That Infer ICP From Intracranial Structures

5.1.1. Ultrasound Time of the Flight Techniques

The majority of patented methods for noninvasive monitoring of ICP are based on an assumption that changes in ICP affect the physical dimensions and/or acoustic properties of the cranial vault or intracranial structures (dura, brain tissue, brain ventricles, and/or intracranial vessels). Dimensions of the cranium or its structures are determined with the ultrasound “time-of-the-flight” technique that measures the transit time of an ultrasound wave and its (potentially multiple) echoes on their path through the cranium and calculates the corresponding distance(s) using known ultrasound propagation velocities in different tissues (e.g. bone, brain, or fluid). Three patented methods derive ICP from the diameter of the cranium that is measured with pulsed phase-locked loop ultrasound transducers . Diameter of the cranium is calculated from the transit time of either the incident ultrasound wave detected with two transducers located on the opposite walls of the cranium , or of the echo wave reflected from the contralateral wall of the cranium and detected with the same transducer that generates the incident wave . Unfortunately, a reproducible quantitative relationship between the diameter of the cranium and ICP could not be established because ICP-induced changes in the cranium diameter are very small compared to the inter-individual variability of skull sizes, shapes and thicknesses . Two patents issued to Kageyama and colleagues teach how ICP can be inferred from the thickness of the dura mater that is estimated from interference echoes of ultrasonic wave . The utility of the method was successfully confirmed on four healthy subjects and four patients with intracranial hypertension , but larger validation studies have never been conducted as the method failed to attract enough interest among clinicians. Michaeli proposed that ICP be inferred from the magnitude and shape of pulsations of the third ventricle synchronous with the cardiac cycle or respiration, where the pulsations are measured along the propagation axis of an ultrasound wave. The method so far has not been independently validated, and the author provides no exact data from which one could estimate the accuracy of the method. However, the discussion in the body of the patent document suggests that the method is able to distinguish among three ranges of ICP (<20, 20-40 and >40mmHg) but cannot provide an exact value of ICP within the range. 

More recently, multivariate methods have been proposed that derive ICP by combining the transit times with measured acoustic impedance, resonant frequency and ultrasound velocity , or with dispersion of the ultrasound wave on its way through the brain parenchyma . Ultrasound ICP monitors based on the latter approach, which are being developed at Vittamed Technologijos (Kaunas, Lithuania) , have showed an impressive agreement with invasively measured ICP, with an average difference of only 2-3 mmHg in a small clinical population .  

The common drawback of all afore described methods is that they measure only relative changes of ICP as referenced to a baseline measurement during which absolute ICP is known, i.e. the ultrasound readouts need to be calibrated on each subject against an invasive measurement. In order to eliminate the need for the invasive measurement, Yost devised a method that extracts pulsatile changes in the ultrasound signal caused by beat-to-beat oscillations of arterial blood pressure, and calibrates the ICP monitor by comparing the amplitude of the extracted waveform to absolute difference between the systolic and diastolic blood pressure (which can be easily measured noninvasively). The method has not been properly validated and it remains unclear whether its performance would remain stable across a wide range of intracranial and arterial blood pressure.

In conclusion, ultrasound ‘time of the flight’ methods for non-invasive ICP monitoring have not been extensively validated and currently the majority of them do not seem to be accurate enough for a routine clinical use. Their original formulations usually do not specify locations for the transducers placement, and do not address how the intentional or accidental use of different locations and/or angles of the transducers will affect the reliability of ICP estimates. It has also remained unexplored how the measure-ments are affected by the presence of intracranial pathologic masses. 

Fig. (2). (A). Schematic example of the time-of-the-flight ultrasound measurement (from ). (B). Vittamed’s system for noninvasive monitoring of ICP and intracranial slow, respiratory and pulse waves (from ).

(blood collections or tumors) on the path of the ultrasound wave, or by brain masses shifts. They are however promising because the ultrasound technology is relatively inexpensive and easy to apply, while the trans-ducers and supporting equipment can be easily made hand-portable Fig. (2B). Time of flight methods are also the only among all patented approaches that allow for continuous monitoring of intracranial pressure. 

5.1.2. Transcranial Doppler Ultrasonography

The other group of ultrasound-based methods for ICP monitoring utilizes transcranial Doppler (TCD) ultrasonography . The TCD measures the velocity of blood flow through the major intracranial vessels by emitting a high frequency (>2MHz) wave from an ultrasound probe and detecting a frequency shift between the incident and reflected wave which directly correlates with the speed of the blood (the so called Doppler effect). The measurement is taken over the regions of the skull with thinner walls (temporal region, back of the head, or through the eye) as the bones strongly attenuate the transmission of the ultrasound at these frequencies Fig. (3). TCD is primarily a technique for diagnosing various intracranial vascular disorders such as emboli, stenosis, or vasospasm , and can be used to identify patients who are at risk of developing cerebral ischemia in early phases of traumatic brain injury or stroke . ICP can be estimated from the TCD measurements because it impedes the blood flow and consequently decreases the velocity of blood flow. Besides the mean velocity, pulsatility index (which is the difference between peak systolic and end diastolic velocity, divided by mean

Fig. (3). Transcranial doppler ultrasonography of the cerebral circulation through the temporal bone. 


flow velocity) , fraction of the cycle in systole and slopes of the TCD waveforms have been correlated with ICP. The estimates are however insufficiently accurate with the margin of error of ±10 - 15 mmHg . Much better results were achieved when TCD was used only for a qualitative assessment of patients into one of the three categories (low, normal and high ICP) .

In spite of the insufficient accuracy of ICP estimates, TCD remains an attractive alternative to the standard (invasive) ICP monitoring because of its ability to detect cerebral ischemia (which may or may not be caused by intracranial hypertension). Other advantages of the method include its relative inexpensiveness, ready availability in most hospitals and many outpatient health care facilities, familiarity of medical specialists with the procedure, and portability of the equipment that would allow for in-field use. The main disadvantage, besides the insufficient accuracy of ICP estimates, is that the method requires a trained and skilled physician to find a correct vessel and meaningfully interpret the results of the measurement. This is not a problem in neurointensive care units, where trained physicians are readily available, but may represent the main limiting factor in less specialized facilities with limited human resources. TCD therefore may not be suitable for triage of patients presenting symptoms of intracranial hypertension at emergency rooms, community clinics, or rural medical centers.

5.1.3. Mechanical / Acoustic Methods

Methods from this group attempt to derive ICP from mechanical properties of the skull bones rather than of the intracranial content. The underlying assumption is similar to that of the ultrasound time of the flight techniques: that the skull is not completely rigid, so that changes in ICP result in a small but measurable skull expansion which creates additional stress within the skull bones and modifies their mechanical properties. Historically, Mick’s method was the first to relate ICP to the mechanical transfer function and resonant frequency of the skull bones. The transfer function is derived by applying a wide-band, low frequency (<100Hz) mechanical excitation at one location on the skull (via a piezo-tranducer or an impact hammer) and comparing its spectrum to that of a signal received at another location on the upper half of skull. It is proposed that the measurement be self-calibrated by obtaining the frequency response spectrum from a point on the base of the skull of the same subject, which is assumed not to be affected by ICP, or alternatively, pre-calibrated on subjects with normal ICP. Other methods from this group vary this basic approach of Mick in different ways. In Sinha’s method resonant frequency of the skull bones is determined first, then a sinusoidal excitation at the resonant frequency is delivered through a piezo-transducer, and ICP is calculated directly from the phase difference between the excitatory signal and response detected with a second transducer. Yost and Cantrell divided the process into two steps. In the first step, changes in the circumference of the cranium are calculated from the phase difference between a sinusoidal excitatory signal, delivered with a piezo-transducer, and the response that is received at a distance with another piezotransducer. In the second step, changes in ICP are calculated as a product of the changes in the cranium circumference and the elasticity constant of the skull that has been determined earlier by causing known changes in ICP while measuring the cranium circumference . The same inventors proposed another approach according to which ICP can be estimated from the measurements of the cranium circumference and the difference between the systolic and diastolic arterial blood pressure (the pulse pressure) using a mathematical relation between the ICP, skull elasticity and pulse pressure derived by Ursino and colleagues .

None of the aforementioned methods has been properly validated in relevant clinical populations, and their accuracy is unknown. One may assume however that it would be comparable to the ultrasound time-of-the-flight methods, and thus insufficient for a routine clinical use.

5.1.4. Magnetic Resonance Imaging (MRI)

The method of Alperin utilizes magnetic resonance imaging (MRI) and makes use of the intracranial pressurevolume relationship Fig. (1B) to derive ICP from changes in intracranial volume that are calculated from arterial inflow, venous outflow and CSF fluid flow between the cranium and the vertebrospinal compartment. The blood flow is calculated from blood velocity, which is proportional to the phase difference between the incident and resonant radiofrequency signals, and cross-sectional areas of the main arterial and venous blood vessels (carotid and vertebral arteries, and jugular veins) which are obtained from static MRI scans. CSF flow and movement of the cervical spinal cord are similarly estimated from the velocity measurements and cross-sectional MRI scans of the cervical spinal cord. The agreement between direct ICP measurements and MRI estimates presented in the patent document was excellent in a baboon and in four patients Tables  1  and  2 from . The method is however expensive and impractical for continuous monitoring or repeated assessment of ICP over time, while the MRI equipment cannot be made portable. 

5.1.5. Electroencephalography (EEG)

A method patented by Rosenfeld is based on electroencephalographic visual evoked potentials (VEP), i.e. electrical brain activity elicited by a flashing light and recorded with few occipital EEG electrodes. ICP is estimated from the latency of the second negative-going wave (N2) of the visual evoked potential, and a table is provided by the inventors that relates the ranges of measured latencies to corresponding ranges of ICP. Reliability of ICP estimates can be improved by averaging a large number (1,000- 10,000) of VEPs which cancels out the background EEG activity, enhancing thereby the VEP waveforms. A variant of this method has been recently investigated by Wu and Ji  who reported a linear relationship between ICP and the latency of the third positive-going wave (P3) of VEPs recorded with high-density electrode arrays and extracted with independent component analysis.  

Accuracy of EEG-derived ICP levels is difficult to estimate rigorously because Rosenfeld’s assessment is only semi-quantitative (ranges of ICP are estimated from EEG rather than exact values) whereas Wu and Ji failed to report standard errors of their EEG-estimates of ICP. The impression of the authors of this article is that the accuracy seems to be low, especially for the range of ICP which is of most interest to clinicians (between 20 and 40mmHg). The main reason for this is the large inter-individual variability of latencies of the components of VEPs, which makes it difficult to discern between physiological long latencies and pathological delays caused by moderate increases in ICP. Other factors to which the low accuracy can be attributed include difficulties in recognizing the target VEP component (due to its low amplitude and/or variable morphology), and lesions (primary, or secondary – ischemic) of the parts of the brain involved in generating the target VEP components.  Continuous ICP monitoring from EEG is theoretically possible but practically difficult to achieve. While EEG can be recorded continuously for 8 - 12 hours before the conductive gel on the electrodes dries off and the electrodes need to be replaced, continuous visual stimulation would certainly be fatiguing for conscious subjects, although it might be feasible in unconscious patients. When the method was invented, EEG was considered both sensitive and cumbersome for use in the dynamic settings of an emergency room, intensive care unit, or neurosurgery department. Recently, however, wireless, portable and field-deployable EEG systems have become available and can be applied with ease by various medical personnel after minimal training. 


5.2.1. Tympanic Membrane Displacement (TMD)

Tympanic membrane displacement (TMD) technique, proposed nearly twenty years ago by Marchbanks  exploits the effect of intracranial pressure on the acoustic reflex, i.e. a reflex contraction of the stapedius and tensor tympani muscles in response to a sound. Normally, vibrations of the tympanic membrane (eardrum) elicited by acoustic stimuli are transmitted through the chain of ossicles (malleus, uncus, and stapes) in the middle ear to the oval window of the cochlea Fig. (4). Vibrations of the footplate of stapes transmit through the oval window to the perilymph, which in turn causes the endolymph, the basilar membrane, and the organ of Corti to vibrate, activating ultimately the acoustic sensor cells, the inner hair cells of the organ of Corti. The transfer function of this complex mechanical system under physiological conditions is modulated by the action of two small muscles of the middle ear, the tensor tympani and stapedius. The tensor tympani arises from the cartilaginous portion of the auditory tube and the osseous canal of the sphenoid and, having sharply bent over the extremity of the septum, attaches to the manubrium of the malleus (hammer); its contraction pulls the malleus medially, away from the tympanic membrane, which tenses the membrane. The stapedius, which emerges from the posterior wall of the tympanic cavity of the middle ear and inserts into the neck of the stapes (stirrup), prevents excess movements
of the stapes by pulling it away from the oval window. The action of either muscle therefore dampens vibrations of the ossicles and reduces the amplitude of transmitted sounds for
up to 20dB. The muscles normally contract in response to vocalization, jawing and loud external sounds, which is accompanied with a small but measurable displacement of the eardrum from its initial position. Because cerebrospinal fluid and perilymph communicate through the cochlear 

Fig. (4). Schematic anatomy of the middle ear. 

aqueduct, an increase in intracranial pressure is directly transmitted to the footplate of the stapes, changing its initial position and affecting thereby the direction and magnitude of
the displacement of the eardrum in response to a sound. 

The displacement can be measured with common tympanometers used for impedance audiometry that are portable and relatively inexpensive and easy to use (particularly the modern, computerized tympanometers with fully automated measurement procedure). Inward displacement (negative peak pressure on audiogram) is suggestive of high, and outward of normal or low ICP . The direction and magnitude of TMD, however, depend not only on the initial position of stapes but also on numerous other factors that affect the acoustic impedance (integrity of the eardrum, condition of the ossicles, patency of the Eustachian tube, pressure and eventual presence of fluid or other masses in the middle ear) or the strength of the acoustic reflex (physiological variability of the reflex threshold, functional integrity of the cochlear and facial nerves, degree of eventual sensory hearing loss). In addition, the assumption that the pressure of perilymph is equal to ICP does not hold if the patency of the cochlear aqueduct is compromised, which is often the case in elderly subjects . Accuracy of TMD estimates of ICP was found to be at the order of ±15mmHg , which is not sufficient for a reliable quantitative assessment of ICP in clinical practice. However, when used for a qualitative assessment of ICP into only three categories (increased, normal and low ICP), TMD showed very high sensitivity and specificity in children with shunted hydrocephalus . TMD may also be useful for serial intrapatient measurements to track the course of relative changes in ICP from one measurement to another. 

An interesting method that involves direct manipulations on the tympanic membrane rather than relying on the acoustic reflex was proposed as one of the embodiments of a US patent by Ragauskas . First, a measurement of the position of the tympanic membrane needs to be obtained while ICP is zero (denoted as the baseline position). Equalization of ICP to the atmospheric pressure according to the inventor can be achieved non-invasively by tilting the head up, or the measurement can be taken during a neurosurgical operation. Later on, ICP can be measured by exerting an external pressure to the tympanic membrane and applying simultaneously the same pressure onto the oval window and inner ear (e.g. through the Eustachian tube) until the eardrum is moved back to the baseline position, which will happen when the exerted external pressure equals ICP. No data is provided in the patent nor is available from other sources that could support the utility of the concept in clinical practice. 

5.2.2. Otoacoustic Emission (OAE)

TMD fails to provide accurate estimates of ICP mostly because the acoustic impedance and its changes due to the acoustic reflex are dominantly determined by the structures and functional properties of the middle ear, and only marginally influenced by changes in ICP. A measurable acoustic phenomenon that originates in the inner ear would, at least in theory, allow for more precise assessment of the pressure of the peri- and endo-lymph, and consequently, of ICP. Otoacoustic emission (OAE), which is a sound generated by subtle oscillations of the endo- and perilymph caused by contractions of the outer hairy cells of the inner ear in response to a loud sound, seems to offer such a possibility. The sound is transmitted to the stapes, and further through the ossicles, to the tympanic membrane from which it can be detected with a sensitive microphone inserted into the ear canal.  

OAE is used in clinical practice to test for hearing deficits in babies and children who are too young to cooperate. The equipment can be made portable, and is relatively easy to use. Two approaches are commonly utilized that increase the unfavorable signal-to-noise ratio and facilitate extraction of the OAE waveform: transient evoked otoacoustic emission (TEOAE) and distortion product otoacoustic emission (DPOAE). A TEOAE system repeatedly delivers a wide band audio burst and listens for the response, which occurs 4-20 milliseconds later. The signal-to-noiseratio (SNR) of the system is enhanced by ensemble averaging of a large number (~ 1000) of responses synchronous to the stimulus, in a way that resembles electroencephalographic event-related potentials (ERP). In contrast, a DPOAE system introduces a pair of primary tones f1 and f2 (f1<f2) and listen to the responses that occur at frequencies mathematically related to the primary freque-ncies, the most prominent being the so called ‘cubic’ distortion product fcdt = 2f1 - f2. In a recent US patent issued to Meyerson and colleagues thought the use of both the TEOAE and DPOAE for measurement of ICP. TEOAE is used first to determine the optimum OAE response frequency, after which the pair of pure tones is deployed in a DPOAE paradigm such that the cubic distortion product frequency (2f1 - f2) equals the optimum response frequency while the ratio of frequencies f2/f1 is set to 5:4, and of intensities I2/Ito 6:5. The inventors also proposed formulae that relate ICP to the intensity or phase of the measured OAE signal, and described how the other physiological signals or behaviors that are known to affect ICP such as small oscil-lations of ICP with each heart beat, respiration, or posture changes, can be used to confirm the validity of the obtained measurements (e.g. the absence of modulation of the measured OAE phase with respiration may indicate occlusion of the cochlear aqueduct, in which case OAE cannot provide any information about ICP). There is little data up to date about the clinical utility or accuracy of otoa-coustic emission as a measure of ICP. A pilot study of Frank and colleagues that evaluated different modalities of OAE in 12 healthy volunteers and 5 patients with implanted ventricular catheters for direct ICP monitoring revealed that increased ICP or conditions known to increase ICP (e.g. posture changes, abdomen compression, coughing) were associated with notable decreases (between -2.1 and -7.9SPL) in intensity of the evoked OAE . All results were however reported only as group averages, and no attempt was made to derive a quantitative one-to-one relation between the OAE intensity and ICP. 

5.2.3. Ocular Measurements 

Eye provides another possible window into the pressure changes in the intracranial compartment thanks to the fact that the space between the optic nerve and its sheath is a continuation of the subarachnoid space, and is consequently filled with cerebrospinal fluid whose pressure is equal to intracranial pressure. Intracranial hypertension will thus manifest in increased diameter of the optic nerve sheath, and will impede the blood flow through the central retinal vein that courses within the sheath, along and in part inside of the optical nerve. The impediment of venous return causes visible changes in the eye fundus (venous engorgement, and papilledema, i.e. swelling and elevation of the optic nerve disc) that can be observed with an ophthalmoscope and have therefore been used by clinicians for more than a century as signs of increased ICP. Quantitative assessment of ICP can be made noninvasively in two different ways: by measuring changes in diameter of the optic nerve sheath with an appropriate technique (ultrasound or MRI), or by using ophthalmodynamometry to determine the pressure in the central retinal vein, which is normally slightly higher (1- 2mmHg) than ICP. 

Intracranial hypertension also induces changes at the cellular or axonal level such as the swelling of the fibers of the optic nerve that form the innermost layer of the retina (so called nerve fiber layer – NFL). The information provided by the classic ophthalmoscopy is however only qualitative and may be inconclusive during early phases of intracranial hypertension since it usually takes between two and four hours from the onset of ICP elevation for a papilledema to develop. A patented method that utilizes optical coherence tomography to measure the thickness of the nerve fiber layer and infers ICP from it laid claims of being able to detect the IH-induced thickening of the retina shortly after the onset of IH, but there has been no data that would support the claims or clarify the relationship between the NFL thickness and levels of ICP. Optic Nerve Sheath Diameter

The use of optic nerve sheath diameter (ONSD) for the assessment of ICP dates back to 1987 when Cennamo and colleagues demonstrated a linear relationship between ICP and the sheath diameter measured with a trans-orbital ultrasound probe in an A-scan mode (principally equivalent to the time-of-the-flight measurements of the cranium diameter). The original measurement method was technically difficult and unreliable because of the nearly coaxial alignment of the optic nerve and propagation axis of the ultrasound wave, but the precision was significantly improved with the use of B-scan (or planar) ultrasound which provided longitudinal cross-section images of the optic nerve and its sheath . Since then, the method has been successsfully validated in several relatively large studies that included patients with severe head trauma , hydrocephalus , intracranial hemorrhage or stroke , liver failure , and climbers with acute mountain sickness . Recently, some authors have used magnetic resonance imaging (MRI) for measuring the ONSD . 

The techniques of ultrasound or MRI measurements of ONSD both belong to the public domain. While the ONSD can at any given point along the optic nerve be measured with a precision of <1mm, reliability of derived ICP levels is plagued by inter-individual variability and the dependance of ONSD magnitude on the point along the nerve at which the measurement was taken. Almost all validation studies so far have recommended that ONSD be used for identification of patients with intracranial hypertension that requires treatment (ICP>20mmHg, i.e. ONSD>5mmHg) rather than for a measurement of ICP. The ultrasound technique is relatively inexpensive and easy to use, and the equipment can be made portable and field-deployable. Continuous monitoring of ICP would be possible in unconscious or very cooperative conscious patients with smart algorithms for image processsing and ONSD measurement, but the technique is more likely to be used for repeated rather than continuous assessment of ICP. However, it is not applicable in the case of an ocular trauma (that often concurs a traumatic brain injury), and will give erroneous ICP estimates in patients with conditions that affect the optic nerve such as inflammation, trauma, or compression by orbital or intracranial masses (cysts or blood collections). MRI scans are routinely made upon admission in many specialized institutions in patients with suspected traumatic brain injury or stroke, but in the context of ONSD measurement they provide little advantage over ultrasound in terms of precision. Additionally, MRI is expensive and requires a dedicated facility and highly skilled personnel, which all makes it unsuitable for continuous or repeated noninvasive assessment of ICP. Althout portable MRI systems have been recently introduced , they are used only for diagnosis of bone and joint lesions on the extremities, and it remains to be determined whether they can be reliably used for ONSD measurements. Ophthalmodynamometry

Ophthalmodynamometry or the measurement of the retinal venous outflow pressure (VOP) is performed by applying external pressure on the sclera, for example with a spring plunger, while observing the retinal vessels through an ophthalmoscope. The pressure is gradually increased until the central retinal vein begins to pulsate, which happens at the point when the applied external pressure nears the VOP and is approximately equal to ICP. The original method was described in 1925 by Baurmann and belongs to the public domain, but several modifications have been recently patented that combine the classic ophthalmodynamometry with reflectance oximetry of the retina or ultrasound measurement of blood flow in the central retinal artery , or automate the method by adding a camera and an image processing software capable of recognizing venous pulsations from a sequence of images of the eye fundus . Evaluation in patients confirmed a strong linear relationship and clinically negligible differences (2-3mmHg) between VOP and the invasively measured ICP . The basic equipment needed for the measurement (a plunger and ophthalmoscop) is light-weight and portable. However, the method is unsuitable for continuous monitoring of ICP and cumbersome if frequent repeated measurements are needed. Ophthalmodynamometry requires dilated pupils, a skilled physician or medic and collaboration of the patient, which all hampers its applicability in the field. It cannot be applied in cases of ocular trauma or conditions that selectively affect the optic nerve, and gives erroneously high readings in the presence of a papilledema, which may persist long after ICP has returned to normal. Finally, application of external pressure may trigger the oculo-cardiac reflex and cause transient hypotension, which may be dangerous in patients with intracranial hypertension as it worsens the brain ischemia. 

5.2.4. Manipulations on the Jugular Vein

The method of Allocca consists of occluding the jugular vein for a short period of time (~ 5 seconds) and measuring non-invasively, with a Hall sensor or an ultrasound transducer, the rate of change of blood flow in the jugular vein upstream of the occlusion. Data is provided in the patent document from experiments on cats that demonstrate a linear relationship between ICP and the rate of change of the jugular blood flow. While the method is simple and can easily be made portable and field-deployable, its clinical utility would be plagued with two problems: ICP estimates are inflated since occlusion of the jugular veins is known to increase ICP, and the additional increase of ICP and obstruction of blood flow caused by the measurement procedure might provoke dangerous complications in patients with already high ICP and borderline or insufficient cerebral perfusion.


In spite of a relatively long history of the field and the existence of a number of interesting approaches the use of noninvasive methods for ICP monitoring is still in the exploratory phase. The main reason for such state of affairs is that none of the methods have been found sufficiently accurate and easy to use at the same time. A glance at Table 4, which summarizes the information laid out in section 5 of the article, reveals that most methods fall far behind the AAMI standards, with the margins of error of ICP estimates being of the same order of magnitude as the whole range of ICP that is clinically of interest (0 - 50mmHg). The noninvasive methods can therefore reliably identify only subjects with low to normal or very high ICP, but not the clinically most important population with moderately increased ICP (15 – 30mmHg). Consequently, the measurements add little or nothing to the old-fashioned reliance on clinical signs of intracranial hypertension, which also allow for a reliable distinction between people with normal and

Table 4. Comparison of Methods for Noninvasive Monitoring of ICP.

very high ICP but are not reliable indicators of mild to moderate IH (15 – 30mmHg), and have therefore had little appeal to clinicians. On the other hand, the most precise method (ophthalmodynamometry) is unsuitable for continuous monitoring of ICP and rather cumbersome for frequent repeated assessments. At present, the ultrasound time-of-the-flight technique seems to be the most promising technology as it offers acceptable accuracy of ICP estimates together with ease of use and the ability of continuous monitoring. Some methods are neither very accurate nor easy to use (e.g. TCD, OAE) and some have not been independently validated on a clinical sample and their precision is unknown. The exact reasons for the insufficient accuracy somewhat vary from one method to another but in most cases include large inter-individual variability of the measured anatomical or physiological characteristic and the fact that the measured characteristic is not only a function of ICP but also of other factors such as arterial blood pressure and preserved cerebral autoregulation (e.g. TCD, multivariate ultrasound methods), functional integrity of particular brain regions (EEG, TMD), presence of intracranial masses (e.g. time of the flight ultrasound techniques, ocular methods), or patency of the cochlear aqueduct (TMD, OAE). The problem of dependence of the measured characteristics on multiple factors besides ICP could be addressed by simultaneously measuring some or all of these factors with a multi-sensory device. For an example, the presence of epidural or subdural blood collections could be detected with a near infrared spectroscopy (NIRS) probe that could be used along with an ultrasound transducer, while arterial blood pressure could be simultaneously recorded from extremities.

Another shortcoming shared by nearly all methods is that they cannot be readily used for continuous monitoring of ICP. Continuous monitoring is possible only with static ultrasound or acoustic measurements, where the signal processsing can be easily automated and probes further miniaturized. EEG and OAE enable continuous  recording, but concurrent visual or auditory stimulation can only be intermittent. Ophthalmodynamometry and tympanic membrane displacement are discrete measurements by construction, while TCD and measurement of blood flow velocity in the jugular vein cannot at present be reliably automated and require an extensive involvement of an experienced physician. However, the opinion of the authors of this article is that, depending on the clinical condition and level of health care, ICP monitoring need not be continuous in order to meet the needs and goals of clinicians. For example, if the goal is to recognize patients with incipient intracranial hypertension at the point of contact, mean ICP measured at regular time intervals over a period of time (e.g. every 5 minutes over one hour) should be sufficient. Disease states in which changes in ICP are chronic (e.g. communicating hydrocephalus) and/or progress relatively slowly (e.g. early malignant MCA stroke, or intraventricular hemorrhage) also do not require continuous ICP monitoring. However, conditions with poor intracranial compliance and impeding herniation syndromes do require continuous ICP monitoring, and ICP thresholds breached for certain period of time (e.g. 20mmHg over 15 minutes, 25mmHg over 10 minutes, or 30mmHg over 5 minutes) are relied on when making therapeutic decisions. 

The level of skills required of the personnel varies considerably from one method to another and depends on the  degree of automation that can be achieved with a particular method. If powerful algorithms are implemented for processing of the static ultrasound waveforms or acoustic measurements of the cranial bones, these methods could be applied by nurses, medics, or even personnel without medical education. Prior training and moderate level of (mostly manual) skills would be required of personnel who would be applying EEG electrodes, tympanometers, or sensitive OAE microphones, assuming again that the signals are processed by smart algorithms that do not rely on human interpretation. High level of skills are required for application of ultrasound measurements of ONSD, while TCD, ophthalmodynamometry and measurement of blood flow profile in an occluded jugular vein can only be performed by an experienced physician. The minimum level of skills required by a method is not of crucial importance if the method will dominantly be used in hospitals where trained staff should be available, but will be a key limiting factor for the use of the method in ambulatory settings or in the field, especially since all methods except for MRI are, or can be easily made portable thanks to advancements in technology.

Besides the patented methods discussed in this article, there are several other methods proposed in the scientific literature for noninvasive monitoring of intracranial pressure, such as the assessment of critical closing pressure with TCD or laser-Doppler flux or noninvasive measurement of cerebral perfusion pressure (CPP) based on TCD . These methods belong to the public domain, and as such, have intentionally not been discussed in details in this review. Their accuracy, technical complexity and ease of use are similar to that of the TCD based methods discussed in Section 5.1.2. Interested readers are encouraged to consult the references for more information. 

The future of the field will crucially depend on substantial improvements of the currently insufficient accuracy and enhancing the ease of use of  noninvasive ICP monitoring. Most promising candidates among the existing methods are the multivariate ultrasound technique of Ragauskas that could be made less obtrusive, and ultrasound optic nerve sheath measurements whose accuracy as well as ease of use could be boosted by smart image processing algorithms that would reduce the human involvement to merely positioning the probe over the eye. Another promising but unexplored approach is the use of multi-sensory devices that would not only monitor ICP but also detect the presence of factors that may confound the ICP measurement, and use this information to correct the ICP estimates. This would be in line with the current focus on multi-modal brain monitoring in the management of patients with acute intracranial hypertension, as experts recognize that reliance on ICP is suboptimal and parameters such as brain tissue oxygenation, cerebral blood flow or cerebral metabolism should be monitored along with ICP. Yet another approach could involve focusing on intracranial veins, whose compliance and subsequently blood flow velocity are strongly affected with ICP, which could be measured with TCD or perhaps with optical methods such as near infrared optical spectroscopy (NIRS).


Financial support, provided by the DARPA award W31- P4Q-09-C-0291, is gratefully acknowledged. Dr. Popovic is grateful to his wife, Karolina, for her assistance in preparation and editing of the figures in the article. 


Dr. Popovic has invented a method for noninvasive monitoring of ICP based on near infrared optical spectroscopy (NIRS) that was submitted to the US Patent Office in January 2009 in a form of provisional application. This method has not been discussed nor cited in the article. The other two authors have no conflict of interest to report.


  1. Smith M. Monitoring intracranial pressure in traumatic brain injury. Anesth Analg 2008; 106: 240-248.

  2. Mokri B. The Monroe-Kellie hypothesis: Application to CSF volume depletion. Neurology 2001; 56(12): 1746-1748.

  3. Zhong J, Dujovny M, Park HK, et al. Advances in ICP monitoring techniques. Neurol Res 2003; 25: 339-350.

  4. Lozier AP, Sciacca RR, Romagnoli MF, Connolly ES Jr. Ventriculostomy-related infections: A critical review of the literature. Neurosurgery 2002; 51: 170-181.

  5. Zabramski JM, Whiting D, Darouiche RO,  et al. Efficacy of antimicrobial impregnated external ventricular drain catheters: A prospective, randomized, controlled trial. J Neurosurg 2003; 98: 725-730.

  6. Citerio G, Andrews PJ. Intracranial pressure. Part two: Clinical applications and technology. Intensive Care Med 2004; 30: 1882- 1885.

  7. Munch E, Weigel R, Schmiedek P, Schurer L. The Camino intracranial pressure device in clinical practice: Reliability, handling characteristics and complications. Acta Neurochir (Wien) 1998; 140: 1113-1119.

  8. Fernandes HM, Bingham K, Chambers IR, Mendelow AD. Clinical evaluation of the Codman microsensor intracranial pressure monitoring system. Acta Neurochir Suppl 1998; 71: 44-46.

  9. Citerio G, Piper I, Cormio M,  et al. Bench test assessment of the new Raumedic Neurovent-P ICP sensor: A technical report by the Brain IT group. Acta Neurochir (Wien) 2004; 136: 1221-1226.

  10. Piper I, Dunn L, Contant C,  et al. Multi-centre assessment of the Spiegelberg compliance monitor: Preliminary results. Acta Neurochir Suppl 2000; 76: 491-494. 

  11. Martinez-Manas RM, Santamarta D, de Campos JM, Ferrer E. Camino intracranial pressure monitor: Prospective study of accuracy and complications. J Neurol Neurosurg Psychiatry 2000; 69: 82-86.

  12. Koskinen LO, Olivecrona M. Clinical experience with the intraparenchymal intracranial pressure monitoring Codman MicroSensor system. Neurosurgery 2005; 56(4): 693-698.

  13. Miyake H, Ohta T, Kajimoto Y, Matsukawa M. A new venticuloperitoneal shunt with a telemetric intracranial pressure sensor: Clinical experience with 94 patients with hydrocephalus. Neurosurgery 1997; 40(5): 931-935. 

  14. Brainin M, Bornstein N, Boysen G, et al. Acute neurological stroke care in Europe: Results of the European stroke care inventory. Eur J Neurol 2000; 7: 5-10.

  15. Rutland-Brown W, Langlois JA, Thomas KE, Xi YL. The incidence of traumatic brain injury in the United States, 2003. J Head Trauma Rehabil 2006; 21(6): 544-548. 

  16. Thurman DJ, Alverson C, Dunn KA, Guerrero J, Sniezek JE. Traumatic brain injury in the United States: A public health perspective. J Head Trauma Rehabil 1999; 14: 602-615.

  17. Marmarou A, Anderson RL, Ward JD,  et al. Impact of ICP instability and hypotension on outcome in patients with severe head trauma. J Neurosurg 1991; 75: S59-66

  18. Cremer OL, van Dijk GW, Brekelmans GJ,  et al. Effect of intracranial pressure monitoring and targeted intensive care on functional outcome after severe head injury. Crit Care Med 2005; 33: 2207-2213.                                                                                                                                                                                                                                                                                            

  19. The Brain Trauma Foundation. The american association of neurological surgeons. The joint section on neurotrauma and critical care. Indications for intracranial pressure monitoring. J Neurotrauma 2007; 24: S37-44.

  20. Bulger EM, Nathens AB, Rivara FP, Moore M, MacKenzie EJ, Jurkovich GJ. Management of severe head injury: Institutional variations in care and effect on outcome. Crit Care Med 2002; 30: 1870-1876.

  21. Fakhry SM, Trask AL, Waller MA, Watts DD. Management of brain-injured patients by an evidence-based medicine protocol improves outcomes and decreases hospital charges. Cent Nerv 2004; 56: 492-499.

  22. Robertson CS, Valadka AB, Hannay HJ,  et al.  Prevention of secondary ischemic insults after severe head injury. Crit Care Med 1999; 27: 2086-2095.

  23. Grande PO, Asgeirsson B, Nordstrom C. Aspects on the cerebral perfusion pressure during therapy of a traumatic head injury. Acta Anaesthesiol Scand Suppl 1997; 110: 36-40.

  24. Stocchetti N, Penny KI, Dearden M,  et al. Intensive care management of head-injured patients in Europe: A survey from the European brain injury consortium. Intensive Care Med 2001; 27: 400-406.

  25. Sahjpaul R, Girotti M. Intracranial pressure monitoring in severe traumatic brain injury-results of a Canadian survey. Can J Neurol Sci 2000; 27: 143-147.

  26. Hesdorffer DC, Ghajar J. Marked improvement in adherence to traumatic brain injury guidelines in United States trauma centers. J Trauma 2007; 63(4): 841-847. 

  27. Robertson CS, Valadka AB, Hannay HJ,  et al. Prevention of secondary ischemic insults after severe head injury. Crit Care Med 1999; 27: 2086-2095.

  28. Yost, W.T., Cantrell, J.H.: US5617873 (1997).

  29. Johnson, R., Quirk, W.H.: US20016231509 (2001).

  30. Yost, W.T., Cantrell, J.H.: US20026413227 (2002).

  31. Petkus V, Ragauskas A, Jurkonis R. Investigation of intracranial media ultrasonic monitoring model. Ultrasonics 2002; 40: 829-833. 

  32. Kageyama, N., Kuchiwaki, H., Ito, J., Sakuma, N., Ogura, Y.: US4984567 (1991).

  33. Kageyama, N., Kuchiwaki, H., Ito, J., Sakuma, N., Ogura, Y., Minimiyama, F.: US4971061 (1990).

  34. Kuchiwaki H, Inao S, Andoh K,  et al. Continuous recording of changes in intracranial pressure using interference echoes from ultrasonic waves: A preliminary report of practicality and clinical evaluation. J Clin Ultrasound 1992; 20(7): 447- 455.

  35. Michaeli, D.: WO00068647 (2000).

  36. Bridger, K., Cooke, A.V., Crowne, F.J., Kuhn, P.M., Lutian, J.J., Passaro, E.J., Sewell, J.M.: US5919144 (1999).

  37. Ragauskas, A., Daubaris, G.: US5388583 (1995).

  38. Vittamed Technologijos company website.

  39. Ragauskas A, Daubaris G, Ragaisis V, Petkus V. Implementation of non-invasive brain physiological monitoring concepts. Med Eng Phys 2003; 25(8): 667- 678.

  40. Yost, W.T., Cantrell, J.H.: US20026457147 (2002).

  41. Yost, W.T., Cantrell, J.H.: US20046761695 (2004).

  42. Alperin, N.: US5993398 (1999). 

  43. Madsen, J.R., Tailor, G.A.: US20006086533 (2000).

  44. Crutchfield, K., Mozayeni, R.B., Davidson, J.B., Fitall, S., Lall, R.: US20067104958 (2006).

  45. Ragauskas, A., Daubaris, G., Dziugys, A.: US5951477 (1999).

  46. Report of the American Academy of Neurology, Therapeutics and Technology Assessment Subcommittee. Assessment: Transcranial Doppler. Neurology 1990; 40: 680-681.

  47. Voulgaris SG, Partheni M, Kaliora H, Haftouras N, Pessach IS, Polyzoidis KS. Early cerebral monitoring using the transcranial Doppler pulsatility index in patients with severe brain trauma. Med Sci Monit 2005; 11(2): CR49-52.

  48. Bellner J, Romner B, Reinstrup P,  et al. Transcranial Doppler sonography pulsatility index (PI) reflects intracranial pressure. Surg Neurol 2004; 62(1): 45-51.

  49. Schmidt B, Czosnyka M, Raabe A,  et al. Adaptive noninvasive assessment of intracranial pressure and cerebral autoregulation. Stroke 2003; 34:84-89. 

  50. Aggarwal S, Brooks DM, Kang Y, Linden PK, Patzer JF. Noninvasive monitoring of cerebral perfusion pressure in patients with acute liver failure using transcranial Doppler ultrasonography. Liver 2008; 14(7): 1048-1057.

  51. Mick, E.: US5074310 (1991).

  52. Mick, E.: US5117835 (1992).

  53. Sinha, D.N.: US20006117089 (2000).

  54. Yost,W.T., Cantrell, J.H.: US20046746410 (2004).

  55. Yost, W.T., Cantrell, J.H.: US20046740048 (2004).

  56. Yost, W.T., Cantrell, J.H.: US20046773407 (2004).

  57. Ursino M. A mathematical study of human intracranial hydrodynamics. Part I: The cerebrospinal fluid pressure. Ann Biomed Eng 1998; 16: 379-401.

  58. Alperin, N.: US5993398 (1999).

  59. Rosenfeld, J.G., Watts, C., York, D.H.: US4564022 (1986).

  60. Wu X, Zi J. Non-invasive detection of high intracranial pressure with FVEP picked up by independent component analysis. J Biomed Eng (Chinese) 2007; 24(5): 1015-1018.


  62. Marchbanks, R.J.: US4841986 (1989).

  63. Wagner N, Walsted A. Postural-induced changes in intracranial pressure evaluated non-invasively using the MMS-10 tympanic displacement analyzer in healthy volunteers. Acta Otolaryngol Suppl 2000; 543: 44-47. 

  64. Phillips AJ, Marchbanks RJ. Effects of posture and age on tympanic membrane displacement measurements. Br J Audiol 1989; 23(4): 279-284.

  65. Marchbanks RJ, Reid A. Cochlear and cerebrospinal fluid pressure: Their inter-relationship and control mechanisms. Br J Audiol 1990; 24(3): 179-187. 

  66. Samuel M, Burge DM, Marchbanks RJ. Quantitative assessment of intracranial pressure by the tympanic membrane displacement audiometric technique in children with shunted hydrocephalus. Eur J Pediatr Surg 1998; 8(4): 200-207. 

  67. Shimbles S, Dodd C, Banister K, Mendelow AD, Chambers IR. Clinical comparison of tympanic membrane displacement with invasive ICP measurements. Acta Neurochir Suppl 2005; 95: 197- 199. 

  68. Shimbles S, Dodd C, Banister K,  et al. Clinical comparison of tympanic membrane displacement with invasive ICP measurements. Physiol Meas 2005; 26:1085-1092.

  69. Ragauskas, A.: US20067147605 (2006).

  70. Meyerson, S.C., Avan, P.A., Buki, B.: US20036589189 (2003).

  71. Frank AM, Alexiou C, Hulin P, Janssen T, Arnold W, Trappe AE. Non-invasive measurement of intracranial pressure changes by otoacoustic emission (OAE) – a report on preliminary data. Zentralbl Neurochir 2006; 61(4): 177-180. 

  72. Borchert, M.S., Lambert, J.L.: WO98034536 (1998).

  73. Cennamo G, Gangemi M, Stella L. The correlation between endocranial pressure and optic nerve diameter: An ultrasonographic study. Ophthalmic Echography 1987; 7:603-606.

  74. Hansen HC, Helmke K. Validation of the optic nerve sheath response to changing cerebrospinal fluid pressure: Ultrasound findings during intrathecal infusion test. J Neurosurg 1997; 87(1): 34-40.

  75. Geeraerts T, Launey Y, Martin L,  et al. Ultrasonography of the optic nerve sheath may be useful for detecting raised intracranial pressure after severe brain injury. Intensive Care Med 2007; 33(10): 1704-1711.

  76. Soldatos T, Karakitsos D, Chatzimichail K, Papathanasiou M, Gouliamos A, Karabinis A. Optic nerve sonography in the diagnostic evaluation of adult brain injury. Crit Care 2008; 12(3): R67.

  77. Goel RS, Goyal NK, Dharap SB, Kumar M, Gore MA. Utility of optic nerve ultrasonography in head injury. Injury 2008; 39(5): 519-524.

  78. Newman WD, Hollman AS, Dutton GN, Carachi R. Measurement of optic nerve sheath diameter by ultrasound: A means of detecting acute raised intracranial pressure in hydrocephalus. Br J Ophthalmol 2002; 86(10):1109-1113.

  79. Geeraerts T, Merceron S, Benhamou D, Vigué B, Duranteau J. Non-invasive assessment of intracranial pressure using ocular sonography in neurocritical care patients. Intensive Care Med 2008; 34(11): 2062-2067. 

  80. Helmke K, Burdelski M, Hansen HC. Detection and monitoring of intracranial pressure dysregulation in liver failure by ultrasound. Transplantation 2000; 70(2): 392- 395.

  81. Sutherland AI, Morris DS, Owen CG, Bron AJ, Roach RC. Optic nerve sheath diameter, intracranial pressure and acute mountain sickness on Mount Everest: A longitudinal cohort study. Br J Sports Med 2008; 42(3): 183-188.

  82. Watanabe A, Kinouchi H, Horikoshi T, Uchida M, Ishigame K. Effect of intracranial pressure on the diameter of the optic nerve sheath. J Neurosurg 2008; 109(2): 255-258.

  83. Pine K. Portable MRI debuts at UCI Medical Center. UCI Healthcare News Releases 2001 (online); from June 13 .

  84. Baurmann M. Über die Entstehung und klinische Bedeutung des Netzhautvenenpulses. Ber Zusammenkunft Dtsch Ophthalmol Ges 1925; 45: 53-59.

  85. Denninghoff, K.R.: US20026390989 (2002). 

  86. Querfurth, H.W.: US20067122007 (2006).

  87. Braxton, E.E.: US20060206037 (2006).

  88. Firsching R, Schutze R, Motschmann M, Behrens-Baumann W. Venous opthalmodynamometry: A noninvasive method for assessment of intracranial pressure. J Neurosurg 2000; 93 (1): 33-36:29. 

  89. Allocca, J.A.: US4204547 (1980).

  90. Thees C, Scholz M, Schaller MDC,  et al. Relationship between intracranial pressure and critical closing pressure in patients with neurotrauma. Anesthesiology 2002; 96(3): 595-599.

  91. Panerai RB. The critical closing pressure of the cerebral circulation. Med Eng Phys 2003; 25(8): 621-632.

  92. Lopes-Magana JA, Richards HK, Radolovich DK  et al. Critical closing pressure: comparison of three methods. J Cereb Blood Flow Metab 2009; 29(5): 987-993.

  93. Lang EW, Lagopoulos J, Griffith J,  et al. Noninvasive ceberovascular autoregulation assessment in traumatic brain injury: validation and utility. J Neurotrauma 2003; 20(1): 69-75.