Background

For much of the human anatomy, surface landmarks have been identified to guide the surgeon in this task. In neurosurgery, despite the contents of the skull being relatively fixed, a lack of clear surface landmarks combined with a greater need for accuracy means that this remains problematic.

Common and urgent operations requiring openings to be made in the skull to locate lesions or structures in the brain are guided by basic traditional methods, taking measurements relative to bony landmarks from 2-dimensional scans and transferring them to the patient.

Neuro-navigation systems are now established as a key piece of equipment for both emergency and elective cranial neurosurgery. Applications include ventricular catheter placement, biopsies, burr hole and craniotomy planning, directing approaches and dissection and assessing the extent of resections.

Existing Solutions

Existing systems take imaging modalities such as CT and MRI scans and generate 3-dimensional models, 'segment' them into different tissue densities, facilitate the generation of 3-dimensional objects and targets, and then use visual or electro-magnetic tracking of instruments to register then track the position of equipment relative to the 3-dimensional model of the patient's head viewed in real-time on an LCD monitor. Hardware usually includes a planning station linked to the local PACs server for acquisition of images and a theatre workstation with touch-screen interface along with a connected camera or EM generator and specific instruments with visual markers.

With the advent of augmented-reality devices such as the Microsoft HoloLens headset, smart phones and tablets, there has been interest globally in exploiting their functionality to develop a new generation of neuro-navigation system...

Microsoft HoloLens & Our Approach

The Microsoft HoloLens allows the wearer to see their environment through tinted glass and can then superimpose 2-dimensional information and 3-dimensional holograms. Combined with built-in spatial recognition, the device can relate holograms to the real environment and interpret hand gestures of the user.

3-dimensional models generated from patient's CT or MRI scans can be converted for rendering as 3-dimensional holograms, complete with intracranial structures and pathologies. When 'registered' to the patient's head the wearer as able to visualise and locate structures to guide surgery.

The advantages of such a solution over an existing neuro-navigation system are numerous:

• Visualisation of the hologram and relevant information is superimposed on the patient maintaining the surgeon's gaze on the surgical field and reducing the hand-eye coordination demands of having to revert back to a separate LCD screen


• Space required in theatre is minimal. Headsets worn by the operating surgeon(s) are all that are required. In comparison existing systems require large free-standing camera stations combined with dedicated LCD monitors and PCs, along with all their associated wiring


• Surgeon interaction with the system is through minimal hand-gestures and voice commands only, removing the need for touch-sensitive non-sterile LCDs


• Selection of registration points without a fixed camera and reference array allows easier registration of patients positioned prone or park-bench


• Costs are likely to be significantly less that with existing solutions thanks to the reduction in specialist hardware required


• A significant reduction in theatre hardware brings a valuable reduction in theatre set-up time