Introduction

Brain computed tomography (CT) is the most-frequently-performed initial diagnostic modality for evaluating intracranial pathology in a neurocritical care unit. However, brain CT is sometimes inadequate for certain settings (e.g., when patients cannot remain stationary or when patients are not suitable for transport because of unstable vital signs). It is also often too difficult to transport critically ill patients who show unstable vital signs for neuroradiological evaluations. In addition, transportation of these unstable patients may carry significant risks to the patient? stability during transport or examination. Furthermore, due to radiation exposure during a CT scan, a brain CT has a limited number of possible repeat investigations. Sonography is a noninvasive, safe imaging modality taking very little time, which clinicians can repeat, as frequently as required, at a patient's bedside.2)5) 

Sonography provides real-time imaging and guides various procedures.5-6) However, studies have not actively tested sonography? ability to assess the adult brain because of poor ultrasound penetration through the thick cranium. The neurosurgical field has validated sonography only in limited groups of patients, such as infants.3)

This study aimed to test the diagnostic validity and clinical usefulness of bedside brain sonography in the evaluation of critically ill patients with surgical cranial defects and to consider the future uses of bedside brain sonography in neurocritical care.

Materials and Methods

We performed bedside brain sonography on twelve prospectively-recruited patients with surgical cranial defects. An experienced radiologist and a neurosurgeon performed the sonography scans. All patients were in the intensive care unit following decompressive craniectomies, due to traumatic brain injury and related intracranial hematoma in seven patients, to subarachnoid hemorrhage in four patients, and to intracerebral/intraventricular hemorrhages in one patient. Using a standard ultrasound machine (Envisor HD, Philips Ultrasound, Andover, MA, USA), we performed a bedside brain sonography through each patient's skull defect. We routinely used two types of ultrasonic probes (the 3.5 MHz convex probe and 7.5MHz linear probe). Each brain sonography took place immediately after each patient's brain CT or magnetic resonance imaging (MRI). We obtained transverse and coronal sonography image planes similar to those in CT or MRI, and we correlated the sonography image with the CT or MR image in each patient, correlating fifteen data pairs of brain sonography and brain CT or MRI. We evaluated the following: 1) structural displacement due to brain swelling; 2) ventricle size; 3) presence and extent of intracranial hemorrhage; 4) presence and extent of cerebral infarct; and 5) blood flow patency and velocity of cerebral arteries in the basal cistern, by means of color and pulsed Doppler sonography.

Results

We evaluated twelve patients via bedside sonography, all of whom underwent decompressive craniectomies due to the following: acute subdural hemorrhage, 3 cases; intracerebral hemorrhage (ICH), 3 cases; subarachnoid hemorrhage (SAH), 3 cases; and cerebral infarction, 3 cases. We performed comparison CT or MRI for all patients.

Bedside brain sonography allowed prompt and repeated intracranial evaluations during the postoperative period without the problems inherent in transporting critically ill patients. We collected fifteen data pairs of brain sonography and either brain CT or MRI. All fifteen data pairs showed good correlations with these two image modalities (brain sonography and CT or MRI).

In all cases, we could visualize the major intracranial anatomical structures, pathologies, and flow hemodynamics of the cerebral arteries well via bedside sonographic assessments. Brain sonography through patients' surgical cranial defects revealed the following: 1) the presence and extent of brain parenchymal hemorrhages in three cases; 2) the presence and extent of cerebral infarctions in three cases; 3) shifts of the midline structures in four cases; 4) enlargement of ventricular systems in three cases; 5) flow velocities of the cerebral arteries in the basal cistern in four cases; and 6) the presence and extent of subdural or subgaleal fluid collection in three cases. We could also evaluate the contralateral hemisphere in each case.

Brain sonography helped us to detect each patient's intracranial pathology after their neurosurgical operation. We could assess cerebral edema severity by the brain's degree of structural displacement and gyral effacement, and we could assess the size of intracranial space-occupying lesions and their anatomical correlations with surrounding structures based on multi-plane and real-time evaluations. In patients with aneurysmal SAH, we could easily monitor the development of hydrocephalus or subdural fluid collection via bedside brain sonography. In one patient with aneurysmal SAH, we could carry out repeated sonographic monitoring of the development of hydrocephalus through a large burr-hole site after cranioplasty. We monitored the development of transtentorial herniation and the progression of cerebral swelling in those patients who developed traumatic brain swelling and/or intractable intracranial hypertension.

Conventional images from brain CT or MRI correlated well with the brain sonographic images we obtained (Figs. 1 and 2). The cerebral infarction and intracerebral hemorrhages we detected in this manner were comparable to their images in patients' brain CT scans. Acute cerebral infarction appeared as a homogenous hyperechoic area (Fig. 1), and acute hematomas appeared as a hyperechoic area (Fig. 2).

Another reason for these brain sonography examinations was positioning an external drainage catheter in the ventricular system. In one patient with malignant cerebral edema, we accurately positioned an external drainage catheter in the small, collapsed, and displaced ventricle for monitoring intracranial pressure and draining cerebrospinal fluid. In one case, sonography revealed entrapped parenchymal fluid in the left parietal lobe and offered a chance to treat this by means of an ultrasound-guided aspiration of the entrapped fluid.

Through color and pulsed Doppler sonography, we were able to measure the cerebral arteries' flow velocities in the basal cistern in four patients with aneurysmal SAH who had a high risk of cerebral vasospasm (Fig. 3). In these patients, Doppler sonography did not reveal a significant increase in blood flow velocity in the intracranial arteries' cisternal segments, and these patients did not show any vasospasm symptoms.

Discussion

Brain CT is the initial diagnostic modality of choice for evaluating neurosurgical patients during the postoperative period.4) However, transportation of critically ill patients who need mechanical ventilation or other supportive care may threaten the safety of these unstable patients.

Sonography can be performed from the bedside in a neurosurgical unit or intensive care unit without transport of patients to the examination room. In addition, sonography allows frequent follow-up examinations without a radiation hazard.2)5) Bedside sonography can be rapidly performed during the clinical deterioration phase in critically ill patients. Physicians can carry out medical interventions at the same time. In CT and MRI, patient immobility during the scan is critical, but even irritable patients can undergo brain sonography. This technique's multi-planar capability, simplicity, and real-time evaluation ability are most advantageous. However, sonography's role in neurosurgical practices remains an unexplored issue. The major limitation of sonographic evaluation of the brain is the poor sonic window, through the cranial bone.5) For these reasons, we used sonography to evaluate patients with surgical cranial defects.

Bedside brain sonography can be used in a wide range of neurocritical care conditions, including head trauma, SAH, and non-traumatic ICH. Brain sonography is a useful method for visualizing simple intracranial anatomical structures and pathology and blood flow of the intracranial arteries-especially in the circle of Willis. Speckle noise is an inherent property of medical ultrasound imaging, and it generally tends to reduce image resolution and contrast, thereby reducing this imaging modality's diagnostic value.1) However, in our experience, brain sonography provides as much information as does brain CT. Although we can assess detailed parenchymal structures via CT or MR, bedside sonography reveals most of the information that neurosurgeons need to know during the postoperative period.

Doppler sonography gives hemodynamic information regarding the cerebral arteries in the basal cistern. This real-time imaging technique can provide a visual examination of the cerebral arteries. In four cases of aneurysmal SAH, we performed sonography to identify any blood flow changes in the cerebral arteries that were suggestive of vasospasm. Flow abnormalities, such as stenosis and vasospasm, are assessable via pulsed Doppler sonography.

Brain sonography may be useful in therapeutic interventions. As in our cases, the positioning of intracranial catheters and aspiration of entrapped fluid are feasible with this technique. Our experiences suggest that brain sonography through a surgical cranial defect can be of significant benefit in everyday neurosurgical practice. Bedside brain sonography through a surgical cranial defect has the potential to become an effective diagnostic tool, comparable to CT, in neurosurgical patients, especially in critically ill patients and restless patients. Our experiences allow us to make practical recommendations for use of bedside brain sonography in the field of neurosurgery. 

Conclusion

Bedside sonography allows rapid and repeatable real-time monitoring of critically ill patients who need prompt neuroradiological evaluations. Because of its noninvasive character and suitability for bedside application, physicians can use brain sonography through a surgical cranial defect for emergency or routine diagnostic procedures and repeat them as often as necessary for routine follow-up during everyday neurosurgical practice.

References

1)   Carmo BS, Prager RW, Gee AH, Berman LH. Speckle detection for 3D ultrasound. Ultrasonics 40:129-32, 2002

2)   Gerriets T, Stolz E, Konig S, Babacan S, Fiss I, Jauss M, et al. Sonographic monitoring of midline shift in space-occupying stroke : an early outcome predictor. Stroke 32:442-7, 2001

3)   Graziani L, Dave R, Desai H, Branca P, Waldroup L, Goldberg B. Ultrasound studies in preterm infants with hydrocephalus. J Pediatr 97 624-30, 1980

4)   Kelley RE, DellaBadia J, Minagar A, Kelley BJ, Brunson R. Neuroimaging of the complications of epilepsy surgery. J Neuroimaging 14:33-41, 2004

5)   Rubenstein JB, Pasto ME, Rifkin MD, Goldberg BB. Real-time neurosonography of the brain through calvarial defects with computed tomographic correlation. J Ultrasound Med 3:443-8, 1984

6)   Unsgaard G, Gronningsaeter A, Ommedal S, Nagelhus Hernes TA. Brain operations guided by real-time two-dimensional ultrasound: new possibilities as a result of improved image quality. Neurosurgery 51:402-11, 2002