Introduction

Primary and secondary brain tumors are often treated with a combination of maximal surgical resection followed by radiation and/or chemotherapy [1]. The ability to completely remove an intraparenchymal tumor depends on many factors, including lesion location, the surgeon’s ability to recognize the tumor, the ability to define tumor margins, and the ability to detect residual tumor following excision [1–4]. In addition, these factors must be identified quickly in order to minimize surgery and anesthesia time. In order to ensure best patient outcomes, real‐time and accurate assessment of brain parenchyma and parenchymal lesions intraoperatively is essential. Real‐time diagnostic imaging is used in concert with neurosurgery in order to accomplish this.

Diagnostic imaging has been used to help guide neurosurgery for many years. The first stereotactic frame designed for human use was built in 1918; however, the stereotactic system using three‐dimensional Cartesian coordinates was not developed until 1947 [5]. Intraoperative computed tomography (CT) and magnetic resonance imaging (MRI) scans are available in some facilities and are highly accurate; however, the expense of such systems limits their availability in veterinary settings [1]. In addition, the use of these machines requires that stereotactic equipment be attached to the patient at all times prior to and during the procedure, which can interfere with the surgical approach. Any movement of the equipment results in inaccurate imaging results. Additionally, there can be significant shifting of lesions and normal brain structures during the surgical procedure as a result of resolving mass effect or intraoperative hemorrhage, which can result in significant errors when comparing the pre‐surgical imaging.

In contrast, intraoperative ultrasound is a reliable, non‐invasive, real‐time tool that can be used to accurately assess tumor volume and residual tumor, without the need for stereotactic equipment attached to the patient and without the confounding effects of changing mass effect during the surgical procedure [1–4, 6]. Human studies have shown that intraoperative brain ultrasound images are comparable, and in some cases superior, to MRI images of tumors and normal anatomic landmarks (Figure 18.1) [7]. Although there are some concerns about the application of transducer pressure on the brain parenchyma during the procedure, to date there is no identified risk of mechanical injury to the brain parenchyma resulting from intraoperative ultrasound use. Because the exam is real‐time, intraoperative complications such as hemorrhage can be immediately appreciated using this imaging modality. A standard ultrasound machine can be used for this purpose and, in addition, the ultrasound machine is portable, so there is no need to purchase a separate machine solely for the intraoperative suite.

Ultrasound is a real‐time imaging modality, so the operator can account for changes to the intraparenchymal structures during the surgical procedure. Ultrasound, however, is highly operator dependent with a somewhat steep learning curve. It has been reported that operators who are familiar with MRI and CT images of the brain but who have not been trained on ultrasound struggle to identify anatomy and lesions during the procedure when comparing to preoperative cross‐sectional images (Figure 18.2) [8, 9]. In cases where an experienced sonographer is not available to assist, a stereotactic approach may be preferred. In addition, ultrasound cannot be used preoperatively to help guide the approach; thus, a multimodality approach is preferred when planning and performing surgical removal of intracranial lesions.

Artifacts in Imaging

Unfortunately, ultrasound is prone to many artifacts which can affect image quality and confuse inexperienced sonographers. Generally, these artifacts result from limitations to the width of the ultrasound beam or differences in the attenuation coefficients of intracranial structures [10]. The most common sources of error in intraoperative ultrasound are brightness errors. These errors result from the inclusion of air bubbles, coagulated blood, or hemostatic agents included in the surgical field, and result in the underlying tissues appearing more hyperechoic (Figure 18.3) [10, 11]. Interestingly, even the temperature of the saline infused into the surgical site has the potential to change the echogenicity of the underlying tissues [11].

The best way for sonographers to limit the introduction of artifact into the intraoperative exam is to choose a transducer which best fits the shape and size of the surgical craniotomy. Doing so will reduce the potential to include unwanted air bubbles into the site, and will limit the amount of sterile saline needed to infuse the surgical site. The sonographer should additionally keep the ultrasound probe as close to horizontal as possible. This minimizes air trapping within the surgical site. An additional strategy to reduce brightness artifact is to continually use time‐gain compensation. This will keep the tissues the same echogenicity regardless of depth and will minimize the potential for overinterpreting artificially hyperechoic normal tissues as neoplastic [11].