Introduction
Imaging is the cornerstone of stereotactic surgery. With the advent of digital sectional imaging and its integration with stereotactic devices, imaging has come to mean computed tomography (CT) or magnetic resonance imaging (MRI) (described extensively in this topic). Specifically, it means preoperative scanning and use of a stereotactic frame or "frameless system.” Imaging technology has become increasingly sophisticated—new frames, faster computers, more versatile software, fancier graphics, and Ethernet connections are all means of improving the manipulation of preoperative images. A scan(s) is obtained before surgery, image fusion is done as needed, the plan is made, and surgery completed. Then, the surgeon waits until a postoperative scan confirms that the surgical goals were obtained.
This paradigm is about to change. The ability to acquire high quality neuroimaging in the OR exists, and is becoming increasingly available. Neu-rosurgeons may resist this technology, citing the supposed expense, the lack of significant added value, etc. But we are not strangers to intraoperative imaging. Stereotactic surgery was first done with this technology, in the form of ventriculography. Now that we can perform modern neuroimaging in the OR there seems little reason not to fashion stereotaxis around this capability.
Of course, disagreement exists regarding intraoperative imaging technology, as is expected with the introduction of any new technology.
Intraoperative MRI
Intraoperative imaging with MRI (iMRI) is being accomplished with systems that meet the needs of the neurosurgical OR in different ways. The ideal iMRI will provide the best possible images, create the least amount of interference with the OR as possible, include a means of stereotactic navigation, and be as versatile as possible.
The first iMRI was built at the Brigham and Women’s Hospital in Boston as a collaborative effort between the Neurosurgery and Radiology departments and General Electric Medical Systems [1]. This system, the Signa SP, employs two vertically oriented magnet poles in a 0.5 T superconducting magnet. The patient is always positioned within the magnet and stereotactic space is defined as such. The Signa SP includes an integrated infrared-based navigational device. A full set of MRI compatible OR equipment was designed for use with this system, which is installed in a specially constructed suite. Other services may use the Signa for such procedures as nasal sinus surgery, breast biopsy and hyperthermic treatments, and prostate implants.
The Siemens corporation developed a 0.2 T iMRI known as the Mag-netom Open [2]. It has 2 horizontally oriented poles with a 25 cm gap and using the "fringe field” concept regular OR instrumentation can be used, with the patient rotated into the magnet for imaging [3]. Positioning options are limited and navigation is not available. Siemens has developed a newer system, with a 1.5 T magnet, that will allow for any patient position and includes an infrared navigational tool (Falhbusch, personal communication, 2002). The fringe field concept will be used and the OR table, mounted on a special pedestal, will be rotated towards the gantry for imaging.
Other investigators have developed 1.5 T iMRI units that have been in use for several years. Sutherland spearheaded a project in Calgary whereby a ceiling mounted device is stored in an alcove adjacent to the OR and moved in on a track for imaging [4,5]. The patient and table are enclosed in a copper shield during image acquisition, so the Calgary system allows for the use of regular OR equipment. It also has an integrated IR navigational tool. Hall and Truwit, at the University of Minnesota, adapted a Philips 1.5 T MRI for shared use—on one side of the bore is an OR, and on the other side patients can be brought in for the full gamut of image acquisitions [6]. On the OR side, most of the surgery can be done far enough from the magnet to allow for the use of regular instrumentation; still, there are limits to positioning and navigation is not available. The authors have developed a very useful skull-mounted device (Navigus, Image-Guided Neurologics, Melbourne, Florida) for stereotactic targeting using MR images acquired at the time of surgery [7].
The shared use concept was also employed at the University of Cincinnati, where a 0.2 T Hitachi MRI with a horizontal magnet gap was adapted for OR use and at other times for diagnostic imaging [8]. Despite the low magnet field strength, routine brain and spine imaging can be done with this system. However, navigation is not integrated and patients must be transported from the main OR to the MR suite for imaging.
A lower magnetic field was employed in the development of the Odin PoleStar iMRI. Hadani and collaborators designed a system using a 0.12 T magnet, built from scratch primarily to enhance intracranial neurosurgery [9]. This device is meant to function in a regular OR; radiofrequency interference is eliminated by shielding the room or more recently, by using a ”local shield” that surrounds the patient during imaging. Scanning and other functions are controlled by the surgeon or the OR staff. The field of view is 16 cm by 11 cm, more than enough to image most intracranial lesions. The magnet poles typically sit under the OR table and are raised when imaging is done. A navigational probe using passive infrared reflecting spheres is integrated in this system, and updating stereotactic accuracy is done simply by acquiring a new scan.
These systems represent the various approaches to iMRI that are available today and whose uses are being investigated. Clearly, there are major conceptual differences between these devices, which also range widely in price from about $750,000 to $4,000,000. A neurosurgical department contemplating implementing iMRI will need to examine carefully what its needs are and, how much they want to spend on installing a new technology. There is some new data that suggest that iMRI guidance yields shorter stays in the ICU and the hospital, which may help to justify purchase prices to some extent [10].
Intraoperative CT
Digital imaging in the neurosurgical OR was first done using CT, where a dedicated unit was installed at the University of Pittsburgh [11]. In general, CT scanners are less expensive than MRI units, but acceptance of this technology has been limited by several factors. These include the lower soft tissue contrast compared to MRI, bone hardening artifact in the posterior fossa, the use of ionizing radiation, and likely need for a dedicated technologist to operate the system. Nonetheless, the rapid scan times and the potential applications for spine surgery make iCT potentially useful, and investigators have continued to evaluate its use [12].
Ultrasound
Ultrasound (US) has been used in neurosurgery for over 20 years [13]. Its major limitation has been image quality and the need to image through the open skull. Neurosurgeons accustomed to interpreting CT and MR images may not readily understand the US images that reflect echogenicity, and items such as cottonoids that have an echo must be removed before imaging. Interpreting the two dimensional images in a 3D manner may also pose problems.
Still, a useful iUS system is attractive because it has a lower price tag than iMRI or iCT. Also, ionizing radiation would not be used, and room shielding or instrument modification would not be needed. As a result, investigation of iUS imaging and navigation technology continues [14].
Conclusion
It is natural to question the utility of intraoperative imaging, especially as the available methods range considerably in price. Some cost benefit needs to be demonstrated, keeping in mind that much of contemporary neurosur-gical technology has become part of the routine without demonstrating cost benefit (the operating microscope, deep brain stimulation, intensive care, and stereotactic frames and frameless systems, for instance).
The logic of imaging in the OR is unavoidable. It is another step towards eliminating guesswork in neurosurgery, the process that began with neurological localization 150 years ago, and that is represented by the concept of stereotaxis itself. Debate regarding the best technology and device is healthy. It is unlikely that one single solution will be found as "best” for intraoperative imaging. As the methods and devices mature, neurosurgeons will be able to choose which system best suits their needs. It is hard to imagine that in the near future stereotactic surgery will be done without the ability to plan, adjust, and confirm based on images acquired in the operating room.
