Navigation in surgery

The first serious experiments to precisely localize specific anatomical structures within the human body can be traced back to the late nineteenth century [1]. Much has changed since, but the main challenge to specifically target an anatomical structure in safer and less invasive ways remained the same. It was only with the advent of medical imaging in connection with the exponential growth of computer processing capabilities that made precise and safe targeting of anatomy a reality. Medical imaging was an important prerequisite to enable navigation. However, pioneering surgeons remain the driving force behind the development of surgical navigation. These clinicians pushed for the development of new technology to solve their surgical challenges. In essence, three key factors pushed the development of navigation in surgery as we know it today: neurosurgery, stereotaxy, and medical imaging.

A surgical navigation system is in some way the same as a commonly used navigation system in a car, for example. Both attempt to localize or determine a position in space in the context of its surroundings. The actual localization technology, however, differs as surgical navigation is not using triangulation like a global positioning system with the help of several geostationary satellites. Modern surgical navigation systems use a stereoscopic camera emitting infrared light which can determine a 3D position of prominent structures, like reflective marker spheres. This allows for real-time tracking of the marker spheres.

For the basic setup, the requirements are a stereoscopic camera, a computer platform with screen, and the respective navigation software. During the surgery, the marker spheres are attached to the patient and at surgical instruments (using reference arrays) to enable an exact localization in space and hence navigation in the operating room (OR).

With each reference array comprising of at least three marker spheres, the computer can calculate the position and orientation of each instrument. A correct localization and virtual display of the instrument on the computer screen is ensured by firmly attaching a reference array to the patient, e.g., in the bone or via a head clamp (see Fig. 1 for a common OR setup in spinal surgery). Movements of the camera intraoperatively are possible because only the relative position of the tracked instruments to the tracked patient reference is relevant.

In the field of neurosurgery and spinal surgery, navigation is usually “image-based,” meaning that imaging data for example preoperative CT or MRI images are required and can be used for navigation in the OR (image acquisition). Before surgery, objects and areas of interest may be planned within the images and hence enrich the data sets (planning). Before the first cut is made, the preoperative image data need to be matched to the current patient position via a registration process (registration). This is the process to establish a relation between the “real” coordinate system as defined by the patient’s reference array and the “virtual” coordinate system of the imaging data. Registration can be paired point-based or use surface matching routines. The surgeon then virtually sees both the current situation and the imaging datasets overlapped and may then navigate on both (navigation). If required, the surgeon may obtain additional images during the surgery and register and include them in the running navigation (intraoperative image acquisition). Figure 2 shows the basic navigation workflow for neurosurgery and spinal surgery.

Modern orthopedic navigation systems are “model-based” and work almost exclusively without information from external image sources. The patients do not need to be exposed to additional radiation, e.g., through CT or X-ray. Instead the navigation software calculates an individual model of the patients’ anatomy based on defined landmarks on the bone which are acquired using a navigated instrument (registration). After an optional planning on the model (planning, e.g., virtual orientation and placement of the joint implant), the actual procedure follows where the surgeon gets supported by relevant information added through the navigation system (navigation). Figure 2 shows the basic orthopedic navigation workflow.

Each surgical discipline and each individual hospital and surgeon have different navigation requirements for their workflow, amount of flexibility, and needed functionality. A wide variety of navigation platforms is available to accommodate all surgical needs: The systems can be installed permanently mounted at the OR ceiling, require minimal OR footprint, or minimize cable clutter and be mobile platforms to be used in several ORs at different times or even be carried around between hospitals for maximum flexibility (see Fig. 3).

The various applications of navigation for neurosurgery, or “neuronavigation,” have been widely reported and published for almost two decades. Neurosurgery was the first surgical discipline to adopt navigation and integrate it successfully in clinical routine. The neurosurgical procedures supported by neuronavigation range widely from intracranial tumor resections to frameless biopsies to pedicle screw placement and stabilizations in the spine. Below is an overview of the main benefits and challenges of neuronavigation in general.

In orthopedic surgery, although each patient is comprised of individual anatomy, the surgeries are quite similar regarding the intraoperative workflow. In each case, a joint must be replaced, with the goal of reproducing nature as well as possible. For both hip and knee replacement, the surgeon aims for precise and accurate placement and alignment of the implants. In contrast to neurosurgery, where the focus lies on precise localization and avoidance of areas of risk, in orthopedic surgery, just a precise measurement tool is required.

There are more than 1.5 million estimated knee procedures per year worldwide. The surgical challenges for total knee replacement are to restore function and to achieve implant longevity. In the cases where a revision of the implant is required, it occurs within 2 years (55 %) or within 5 years (63 %) after the primary total knee replacement. The main causes of this early revision are instability (21–27 %), polyethylene wear and aseptic loosening (10–28 %), and patellofemoral problems (8 %) [17, 18].

The key challenges during hip replacement surgery are the restoration of length and offset of the leg as well as the accurate positioning of the implant. Leg length discrepancy is the second and total hip dislocation is the fifth most cited source of medical malpractice litigation among American Association of Hip and Knee Surgeons [19, 20].

The aim of navigation is to make joint replacement surgery more accurate and more reproducible, not only in highly specialized institutions, but also at average care facilities where 80 % of all joints are implanted. To meet the expectance criteria on the user side, the navigations system needs to seamlessly integrate into the conventional surgical workflow and add little extra cost and effort.

Image-guided navigation yields clinical examples of treatments that only become practical through the benefits of computer-centric patient data, image, and sometimes video-management (e.g., microscope–video integration). A further and natural evolution surrounding navigation systems over the last decade on the high-end has been the digital ORs of neurosurgery suites. These are centered on an image-guided surgery system and integrated with an intraoperative modality. Here, in the treatment of a brain tumor, e.g., an intraoperative MRI is acquired to visualize the progress of a tumor resection. This additional intraoperative information is critical and can be highly beneficial for the remainder of the surgical treatment and the resulting patient outcome [37, 38]. However, to make this feasible, a large amount of data has to be managed adequately in the OR during the procedure—several preoperative scans (e.g., CT and MR), possibly enriched with preoperative planning information—and the intraoperative datasets which undergo further data enrichment from intelligent algorithms like image fusion are automatically registered with the integrated surgical navigation system. In such cases, the additional information becoming available by active and intelligent data management in the OR can be just as relevant as the surgical navigation (transforming the scan data into clinically relevant information by putting it into the correct clinical context, e.g., by image fusion with a preoperative dataset, for example).

It is important to note that such neurosurgery suites with intraoperative imaging are only one of the highest end incarnations of what is generally referred to as “integrated” (and sometimes also touted as “digital”) OR. Generally, integrated ORs have been around for the last two decades [39]. Initially, the key motivation in their creation was to provide additional value to the surgeon and OR staff by making endoscopy video and control more accessible and to enhance these with adjacent functionalities—with the goal of optimizing space, efficiency, and intraoperative decision-making. Thus, integration aspects in ORs to this day still mostly center around central control of endoscope video and parameters, accessed on a joint control panel or touch screen together with surgical equipment like tables, lights, etc. Together, this summarized medical device integration. Only lately, the term “digital OR” has been increasingly used along with an increased focus on digital handling of elements such as endoscope video. Typically, digital video components are used to make a video available on various displays in the OR and allow for enhanced setup flexibility and better OR staff coordination. In addition, streaming or recording of video can be done for teaching purposes in an auditorium, for example, or for documentation purposes. Independent from video management, a typical OR today often comprises a special OR-suitable workstation perhaps wall-mounted with special hygienic measures implemented for the keyboard and mouse user interfaces. This computer is typically intended to access preoperative digital medical image data, such as X-ray or CT scans, which are stored in the facilities’ Picture Archiving and Communication System (PACS), and provide an interface to the Hospital Information System (HIS) by running SAP-client software—summarized as IT access systems.

Thus, in the modern OR, the three realms of medical device integration, image and video distribution and IT access systems are often segregated, as they have their origin in different domains. However, this comes at significant cost: for example (1) hospitals are being locked into OEM-specific solutions for medical device integration and/or video distribution sides which are hard or impossible to alter and expand at a later point; (2) intraoperative image and video data are typically unrelated to patient-specific information making systematic documentation, for example, often a manual and tedious process; and (3) several inconsistent human–machine interfaces inhibit users exploiting available functionalities to their full potential (see Fig. 9).

Nowadays, computer and IP networks can be utilized to tear down existing barriers between medical devices, video distribution, and IT systems, thus enabling better access and improved management of information to facilitate and maximize productivity in the operating room—both for the benefit of the user and patient as well as for the financial benefit of the hospital.

As an example, these benefits become especially apparent when trying to maximize the usefulness of preoperative radiology images. With the widespread adoption of PACS driven by radiology, preoperative medical images had been separated further from the surgeon and staff in the OR: What was formerly available on a physical film for direct interaction in the OR, is now often only accessible via a computer system through a more or less OR-suitable viewing software. Often, there is little to no optimization available for usage in the OR, and only recently, have truly digital ORs emerged, which are optimized for the navigation of information to support on-the-fly clinical decision-making. The widespread adoption of products like Digital Lightbox® from Brainlab [40] show that it can be very beneficial to make preoperative image data available in a simple and intuitive way in the OR—such that the surgeon can ideally immerse in and interact with the data available and is not hindered by the technological shortcomings of an office-viewing solution brought to the OR environment. Furthermore, additional intelligent algorithms such as 3D volume rendering [41] or image fusion [42] can also be deployed to intuitively enrich the raw medical image data, such that critical information becomes more readily available to the surgeon in the OR compared to the usual 2D slice view representations of large 3D CT/MR datasets.

While similar intelligent assistance would also be desirable in real-time for any surgical videos available in an OR, this is technologically far more challenging due to dynamic real-time requirements. Only for certain disciplines do special solutions exist nowadays, for example, in the form of the image-guided surgery systems discussed above.

Furthermore, a truly digital OR also means that all image and video data are fully available for further computer processing and, during this processing, can always be united with any data such as patient demographics retrieved from a connected HIS. Beneficiaries can be intelligent algorithms which further enrich intraoperative image/video data thereby also taking into account initial diagnosis information. Already implemented today is the proper storage of documentation such as screenshots or recordings into digital archives which are being filed automatically under the correct patient ID, etc. An exemplary computer- and IP-centric product is Buzz by Brainlab in which all information processed is fully computer-integrated and thus automatically linked to patient demographics. Additionally, only in such computer-driven information systems can any access to patient-related image and video data including live video streams from the OR be restricted according to state-of-the-art privacy standards, as defined by the Health Insurance Portability and Accountability Act (HIPAA). HIPAA defines the national standards for electronic health care transactions and national identifiers for providers, health insurance plans, and employers in the USA.

The concept of a “digital OR” works quite well for the local navigation of information. But often surgeons and other medical professionals need a more global interconnection. Medical information (e.g., digital medical images), naturally, may not be widely shared, as it is sensitive and personal information.

In days of the aforementioned privacy standards like HIPAA, a secure network needs to be installed, for example, before images may be passed online from an external radiologist to a hospital. Brainlab has developed a clinical online network called Quentry™.