3-D neurosurgical planning tools have been available since the 1980s. The combination of difficult and non-digitized image transfer, manual segmentation, slow computer speed and high cost historically made these systems user unfriendly , and resulted in little popularity with the neurosurgical practitioner. The Dextroscope® first emerged in the mid-1990s in Singapore, with the first case series illustrating the technology being published in 2000. At that time, the editorials commented that the system was highly labor-intensive and, as a result of all the loaded data, required expensive hardware and was at times slow to interact. Since that time, however, there has been a marked development in the power and speed of graphics processors, making the system much faster and more flexible and intuitive. In addition, X-ray technologists have become familiar with 3-D postprocessing from other systems and can apply these skills readily to this system.The labor-intensive aspect of image loading and segmentation can therefore be transferred to a technologist who can prepare the case for the neurosurgeon. All these factors add up to make the Dextroscope a neurosurgeon-friendly tool, without the time and work, yet allowing an extremely clear understanding of the patient s microanatomy. This, in our experience, increases the confidence of the neurosurgeon to handle difficult cases, and also may improve the outcome of the surgery through possible changes in the surgical approach arising from the increase in accurate 3-D information available to the neurosurgeon.
Although pioneering attempts faced long processing times and were restricted to CT data, they were found to be useful in enhancing the surgeon s 3-D perception. The 3-D reconstructions were presented as static screen photographs from different angles. Since then, 3-D image reconstructions built from CT, magnetic resonance angiography (MRA) and CT angiography (CTA) image sequences have been successfully used to plan craniofacial surgery and vascular neurosurgery.
Unfortunately, in most 3-D neurosurgical planning systems the neurosurgeon must interact with the econstructed 3-D image in an awkward, nonintuitive 2-D way by moving a screen-bound cursor with a mouse. This is a considerable restriction given the complex manipulations that the surgeon must perform on the 3-D image for surgery planning. Simulating the removal of bone to access a surgical target or manipulating the image to simulate surgical views, for example, can be tortuous using a mouse and keyboard.
The Dextroscope overcomes this problem by providing an advanced virtual reality environment that allows the neurosurgeon to intuitively interact with and manipulate the 3-D image, using their hands in a similar manner to how they would manipulate a real object. Using stereoscopic visualization technology, the surgeon sees the 3-D image floating in front of them within easy reach and can use flexible 3-D hand movements to rotate and manipulate the image, segment organs and structures, make accurate 3-D measurements, and so on.
Apart from being much faster to work this way than using a mouse and keyboard, this approach also provides the surgeon with a greater degree of control over the 3-D image, with their hands literally being able to reach inside to manipulate the image interior.The 3-D image is no longer a passive image, but an active 3-D virtual representation of the patient’s anatomy.
This high degree of interactivity and control is invaluable when working with 3-D images reconstructed from multi-modal data (where bone, tissue, and vasculature are acquired differently) and image sequences (such as those that track a tumor over time). In these cases, correct interpretation of the structures within the 3-D image is particularly difficult and requires an interactive investigation of the image rather than passively looking at static views.
For neurosurgeons to incorporate this technique into their daily schedules, the virtual reality environment must also be easy and comfortable to use and support shared sessions where many medical experts work together to discuss the same patient.The Dextroscope provides a robust collaborative environment in which the user works in a normal seated position (see Figure 1). A second monitor allows observers to easily view the virtual reality environment and see exactly what the user sees.
The Dextroscope stereoscopically displays its 3-D images on a high-definition monitor; they are then reflected via a mirror into the user’s eyes.Wearing only a pair of liquid crystal display (LCD) shutter glasses— synchronized with the time-split display—the user looks into the mirror to see a stereoscopic image that seems to float behind the mirror.When reaching with both hands behind the mirror, the user experiences the sensation of having their hands in the same workspace as the 3-D object (see Figure 2).The fact that object and hand movements take place in the same apparent position allows for careful, dexterous work.
In one hand the user holds an ergonomically shaped handle with a switch that, when pressed, allows the 3-D image to be moved freely as if it were an object held in real space.The other hand holds a pencil-shaped stylus that is used to select tools from a virtual control panel and perform detailed manipulations and operations on the 3-D image.
The Dextroscope reconstructs 3-D images from any tomographic data, including CT,MRI,MRA,magnetic resonance venogram (MRV), functional MRI and CTA, positron emission tomography (PET), single positron emission computed tomography (SPECT) and, in the near future, DTI. It can work with singlemodality data or any multi-modality combination. Importantly, as long as the tomographic data are in Digital Images and Communications in Medicine (DICOM) format, the 3-D reconstruction is fully automatic and, effectively, instantaneous.
Once the surgeon is in the virtual reality environment, they have access to an extensive set of virtual tools that can be used to manipulate the 3-D image, for example to perform data segmentation to extract surgically relevant structures like the cortex or a tumor; make accurate 3-D measurements along linear and curving structures such as blood vessels; adjust the color and transparency of displayed structures to see deep inside the patient; and even simulate some surgical procedures, such as the removal of bone using a simulated skull drilling tool.
Typical structures that can be segmented are tumors, brain cortex, the ventricles, blood vessels, aneurysms, parts of the skull base, and organs (see Figure 3). Segmentation is undertaken either automatically (when the structures are demarcated clearly by their outstanding image intensity) or through user interaction (using, for example, an outlining tool to define the extent of the structure manually). A virtual ‘pick’ tool allows the user to pick a segmented object and uncouple it from its surroundings for closer inspection. A measurement tool provides accurate measurement of straight and curving 3-D structures within the image, such as vessel diameters or tumor size, as well as angles, such as those between vessels.
Experience with the Dextroscope
The Dextroscope has been used extensively for neurosurgical planning at the Neurosurgical Clinic in Mainz (Germany), at the department of neurosurgery of the Hospital Clinic de Barcelona (Spain), and at the National Neuroscience Institute in Singapore. In the US, experience is now being gained at the University of Pennsylvania and at the University of Medicine and Dentistry of New Jersey in Newark.
Recently, the authors have experienced cases where they changed their initial intra-operative strategy after a planning session with the Dextroscope. This technology is also used for training purposes, as well as for patient education. For larger presentations (e.g. conferences and lectures), there is a projection-based virtual reality interactive visualization system (DextroBeam®). This projection-based system allows the whole audience to view actual three-dimensionally reconstructed data from the patient even as the presenter/lecturer is working with the virtual brain.