May 01, 2014
16 min read

Role of intraoperative imaging in spine surgery continues to expand

You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

We were unable to process your request. Please try again later. If you continue to have this issue please contact

The role of intraoperative imaging during spine surgery has increased in prevalence and capability. Surgeons can now use intraoperative imaging by way of fluoroscopy, three-dimensional imaging, CT and robotics for greater precision, quality assurance and improved patient safety.

“These image guidance systems can be used for many spine surgeries and assist with placement of instrumentation in the thoracolumbar or upper cervical spine,” A. Jay Khanna, MD, MBA, Professor of Orthopaedic Surgery and Biomedical Engineering at the Johns Hopkins University, said. “They are also used for minimally invasive spine surgery procedures.” 

Imaging developments are underway to allow surgeons to visualize more levels during deformity surgery, A. Jay Khanna, MD, MBA, chair of the North American Spine Society radiology section, said.  

Imaging developments are underway to allow surgeons to visualize more levels during deformity surgery, A. Jay Khanna, MD, MBA, chair of the North American Spine Society radiology section, said.

Image: Khanna AJ

“We are attempting to use these modalities in all aspects of our spinal care for patients in the operating room anywhere from the skull all the way down to the tailbone,” Terrence T. Kim, MD, an orthopedic surgeon at Cedars-Sinai Medical Center and director of research and education for the Spine Foundation, told Spine Surgery Today. “More importantly, we are using it in all types of spinal surgeries, ranging from deformity or scoliosis surgery to tumor surgery to degenerative spinal instrumentation.”

“I would say there has been maturation in the sophistication of the imaging,” David W. Polly Jr., MD, professor of orthopedic surgery and neurosurgery and chief of spine surgery at the University of Minnesota, said.

David W Polly Jr

David W. Polly Jr.

According to Polly, the transition in intraoperative spine imaging began with the placement of thoracic pedicle screws, which he started doing in the late 1990s. “At that time, we were doing it primarily based on anatomy and fluoroscopy,” he said. 

The conventional approach for placing screws in the spine has been to use anatomic landmarks, according to Khanna. Looking at the surface of the spine, surgeons determine where screws should go in and at what angle based on their knowledge of the anatomy.

“Some people now use conventional fluoroscopy in the operating room (OR) to locate the starting points for screws, for example, and advance the screws from there; more active image guidance is the next step,” Khanna told Spine Surgery Today .

“We started out with two-dimensional fluoroscopy and then three-dimensional fluoroscopy, and then computed tomography married fluoroscopy and those have all by and large gone by the wayside,” said Richard Fessler, MD, PhD, professor at Rush University Medical School. “Now we are looking primarily at CT-based image guidance.”

In the past, surgeons used CT-based image guidance systems that involved sending a patient for a preoperative CT scan and then linking the CT scan with fluoroscopy images taken of the patient in the OR. Surgeons then used the preoperative CT scans to guide placement.

“This system has gone by the wayside because the position of the patient on the CT scanner is different from the position of the patient on the OR table and because the preoperative CT is no longer accurate once the procedure is underway and the anatomy is altered,” Khanna said. “It requires more frequent registration of the points, which often includes attaching a clamp to the patient.”

Polly said the so-called modified fluoroscopy techniques, which involved processing information from multiple images of the patient into 3-D images, were also unsuccessful. “That is sort of the early version of the intraoperative CT; it was not a sufficient image quality to make most of us happy with the imaging, and it was also expensive,” he said. 

Some surgeons also tried performing procedures in the CT suite using conventional CT scanners.

“The challenge with the conventional CT scanners is you had to reconfigure what you were doing and your equipment to utilize the CT scanner,” Polly said. “The biggest advantage now with the mobile CT scanners used in the OR is they adapt to what we are doing, as opposed to us having to adapt to what they are doing.”


Fluoroscopy vs. newer technology

Fluoroscopy is the most common intraoperative imaging technique and, according to Charles A. Sansur, MD, director of spine surgery at the University of Maryland School of Medicine, it remains the gold standard.

Charles A Sansur

Charles A. Sansur

“I don’t think fluoroscopy is ever going to die,” Sansur said. “You can get instant images without any concern about a lack of correlation between the image and the reference point. The only way these newer intraoperative imaging techniques work is by having an appropriate reference point, and your imaging is only as good as the stability of your reference point.”

Fessler also uses routine fluoroscopy. “Fluoroscopy is fast and one can do one’s surgery without any additional anesthesia time to the patient,” he said. “The disadvantage is exposure to radiation to the surgeon over the course of years, especially when young and learning.”

Khanna and his colleagues now use the Brainlab image guidance system, which is an intraoperative 3-D guidance system. Other systems available for intraoperative spine imaging include CT-guided systems like the O-Arm from Medtronic, the Brainlab Airo system, the Artis zeego system from Siemens and robotic systems like the Mazor Robotics Renaissance Guidance System for placement of pedicle screws.

Polly uses intraoperative 3-D imaging CT, the O-Arm and the Stealth image-guided technology for pedicle screw placement.

“The O-Arm gives you the ability to scan the patient prior to your incision and use that information during the surgery without the need to have the unit in the room,” Sansur said. “Even though the advance is there, I don’t think it is available at a lot of institutions, but I do think that ultimately, it will be a common thing.”

The Brainlab technology provides a sharp, clear image and has satisfactory accuracy beyond four or five levels.

“The only limitation to this technology is you have to have a separate dedicated OR table,” according to Sansur. “The O-Arm can be applied to any standard spine operating table.”

At Cedars-Sinai Medical Center, surgeons also use 3-D intraoperative imaging for spine surgery, in particular intraoperative CT scans — the O-Arm and the Brainlab Airo — to visualize the deep bony tissue, Kim said.

Increased precision

Intraoperative imaging systems help ensure precision, according to sources for this article.

“One of the big issues of spine surgery is we are trying to place screws into narrow corridors in the vertebrae. If we go too far medial, we hit the spinal cord and the nerve roots; if we go too far anterior, we could hit major blood vessels,” Khanna said. “We have to get the screw in the exact right place. The other cases are upper cervical spine cases, where we are doing instrumentation at C1 and C2 and those areas for the screws are also small. The stakes are high.”

Polly and his colleagues recently completed a preliminary study, which evaluated their per-screw placement time.

“We have been as fast as less than a minute per screw and our average time is about 5 minutes per screw,” Polly said. “That includes the acquisition of the intraoperative CT scan to do the navigation and then an intraoperative-check CT to make sure the screws are OK after we have placed them.”

Reporting on their first 2,500 screws using a mobile intraoperative CT imaging system, Polly and colleagues noted a 0% return-to-OR rate for implant malposition.

“The literature would suggest that between 1% and 4% of patients who have screws placed have a need to return to the OR for implant malposition,” Polly said.

He and his colleagues also have evaluated the technology’s use in more challenging cases such as congenital scoliosis.


“Up to 18% of the time, there were pedicles that didn’t exist and so no screws could be placed regardless of whether or not we had image-guided technology,” he said. “That wasn’t always evident on the preoperative imaging studies.”

In a recent meta-analysis, Shin and colleagues found data favored the use of guided imagery in the placement of pedicle screws, with an overall 6% risk for pedicle screw perforation with navigation vs. 15% with conventional pedicle screw insertion. They also found no related neurosurgical complications with image-guided screw placement.

Kim and colleagues recently published data showing a 96.6% accuracy rate with CT navigation for minimally invasive percutaneous posterior spine fusion.

“Our accuracy rates are significantly better than with standard fluoroscopic imaging,” Kim said.

He added: “The reason this technology continues to take off is because our outcome studies are telling us this is not a fly-by-night technology. It is definitely here to stay and it is proving to be superior to the standard way.”

Surgical skill and training

The use of intraoperative imaging technology has also enhanced surgeons’ skill because of the feedback provided by the technology. “When I teach surgeons in training, I can safely let them do more and they get it a lot faster than when I was training surgeons before this technology existed,” Polly said.

“Particularly early in one’s practicing career, that type of image guidance is helpful,” Fessler said. “We have also seen though that to an experienced surgeon, it is not much more accurate than what he does with the fluoroscope.”

Richard Fessler

However, surgeons in training should be cautious about relying too much on intraoperative imaging. “It could become a crutch you are not able to let go of and may compromise your ability to troubleshoot your own cases when things go wrong,” Sansur said. “At a training institution, it would be important to do cases both with navigation and without,” he noted.

Reducing radiation exposure

One of the goals of the current imaging modalities is to decrease the radiation exposure associated with conventional fluoroscopy. “Most people don’t realize one minute of fluoroscopy is equal to 150 chest X-rays,” Khanna said. “Exposure to radiation can lead to cataracts, cancer, cytogenic effects, hematologic effects and skin problems. Surgeons, especially older surgeons before radiation exposure was discussed and studied as much, suffered from these medical conditions.”

With the shift from open spine surgery to minimally invasive surgery, surgeons are performing more procedures that may increase their risk for radiation exposure. “The risk to the patient is not as significant because unlike the patient, the surgeon is in the OR many times a week all year long so his or her cumulative exposure add up in the long-term,” Khanna said. “Without the frequent use of fluoroscopy, radiation can have a significant impact on surgeons and their staff.”

However, according to Fessler, with experience, the radiation exposure with fluoroscopy is significantly reduced because surgeons can complete the procedures faster. 

Because there is still radiation exposure for the patient, “we are trying to find a threshold where we drop the radiation as low as possible and still get an acceptable picture,” Kim said. “We are trying to find that balance, and once we do, I think we’ll find this radiation issue is markedly minimized.”

Newer intraoperative imaging

As with any new technology, the newer imaging systems are expensive. “The issue becomes how much you are willing to spend to improve your accuracy,” Polly said. 

Kim and colleagues at Cedars-Sinai are working on cost-analysis studies of intraoperative imaging systems. “This is an important question that warrants further studies. We are looking at the question of, given the technology is a large hospital capital expenditure, is it cost-effective because of the negative events it prevents or avoids?” he said. “Could this system that is more accurate potentially prevent taking a patient back for a misplaced screw and offset the actual cost of the technology itself?”


Intraoperative imaging also comes with a learning curve, according to Khanna.

“When you start using these systems, it takes a while for the surgeon to learn how to use them,” he said. “But after you learn how to use them, it does save time because you know exactly where to go.”

To alleviate the learning curve at Polly’s institution, surgeons chose to implement the technology on cases in which intraoperative imaging was not necessary. Therefore, when it was needed, it worked well, Polly said. “There were some institutional challenges with this big machine that initially scared people, but it has become so routine now.”

Other disadvantages include disruption in workflow when bringing the image guidance system in and out of the field as well as the potential risk for infection. “You have yet another instrument you are introducing into the field,” Khanna said.

In addition, with current intraoperative imaging systems, the field of view is limited. “If the spine is curved and deformed and I use image guidance, I can only see about three levels at a time,” Khanna said. “That is one of the downsides of the current-generation systems. We are working on some of these limitations.”

Dependence on a separately placed tracking array is another disadvantage of newer systems. “If the array is loosely placed or your assistant bumps it, it gives you completely false information and you may think you are operating in a location where in fact you are not, which would increase the risk for injury to the patient,” Fessler said.

Because of these disadvantages, surgeons like Sansur believe intraoperative imaging technology should be used selectively, such as for patients who have had multiple previous surgeries in whom normal anatomical landmarks are not as easily discernible. Sansur said it would also be appropriate for patients with scoliosis and during thoracic disc removal. 

Kim agreed. “It is all about being responsible and trying to critically analyze and determine when it is the right technology to use. Sometimes patients do not need CT navigation for all their spine surgery. Sometimes, it can be done safely and best without CT navigation.”

In development

Intraoperative imaging technology is constantly being further developed. “We are basically fine-tuning something that has been established as working effectively,” Kim said. “We have been working on getting even clearer pictures on the CT scan. Companies have been developing new scanners that are able to image more of the body at once, whereas currently the technology is at one small defined field.”

There are some efforts to combine CT and MRI to achieve a full-colored picture of the bone and soft tissue in one image. “That is kind of the Holy Grail of where we are trying to go with providing the surgeon with a full view of the spine by melding those two image modalities together,” Kim said. 

Image-guided navigation is also fueling robotic surgery.

Jeffrey Siewerdsen

Jeff Siewerdsen

“Robotic surgery hasn’t really taken off in spine surgery because you can’t tell a robot where to go if you in fact don’t know where you’re going,” Kim said. “Imaging provides an avenue into the implementation of artificial intelligence in the operative setting.”

At the I-STAR Laboratory in the Department of Biomedical Engineering at Johns Hopkins, Khanna collaborates with scientists to create new imaging techniques that aim to improve surgical precision, provide quality assurance and increase patient safety. 

“The collaboration between engineers and surgeons is really vital to advancing the field,” Jeff Siewerdsen, PhD, FAAPM, professor of biomedical engineering in the I-STAR Laboratory at Johns Hopkins, told Spine Surgery Today . “Surgeons know the real clinical challenges first-hand and when they are able to team with engineers, it gives a productive combination. Similarly, biomedical engineers who are driven to learn and understand those key challenges are essential, and together good teams of surgeons and engineers can produce solutions that translate back to the OR in a way that is useful and effective.” 


The I-STAR Laboratory has worked to develop multi-modality imaging techniques, including CT, MRI, ultrasound, fluoroscopy and nuclear medicine, in a form that is well suited to the OR. They have also established technology for overlaying fluoroscopy or video with a hybrid view of images and planning data beyond what is visible in the surgical scene, according to Siewerdsen. “We are able to take a single fluoroscopic frame from a C-arm or a video scene from an endoscope and augment it with 3-D image data,” he said. “A good example is the LevelCheck system that automatically registers information from preoperative CT directly to fluoroscopy as a means of decision support and an independent check in localizing the target vertebral levels in spine surgery. That could improve workflow, reduce stress in making sure the correct level is targeted, and help avoid wrong-level surgery.”

Also underway at I-STAR are ways to improve the quality of intraoperative CT and reduce dose.

Some newer techniques under development at the lab include “higher quality 3-D imaging, better ways of navigating, automatic methods of registering the patient, avoiding the ‘line of sight’ limitation in conventional trackers and development of systems that integrate fluoroscopy with multi-modality imaging,” Khanna said.

Terrence Kim

Terrence T. Kim

“Instead of providing a millimeter accuracy, these new techniques will provide submillimeter accuracy,” he said.

“We are also developing ways in which the imaging itself provides surgical navigation without tracking systems,” Siewerdsen said. “The C-arm and patient provide all the information needed for surgical navigation without additional trackers or markers.”

Going forward, surgeons need to consider at what cost and how much radiation exposure is worth in achieving 100% accuracy every time, according to Polly. “The goal is to continue to figure out how to make it easier to use this technology, to decrease radiation and then hopefully at some point, decrease expenses as well,” he said. – by Tina DiMarcantonio


Kim TT. Neurosurg Focus. 2014; doi:10.3171/2014.1.FOCUS13531.

Larson AN. Spine. 2012; doi:10.1097/BRS.0b013e31822a2e0a.

Larson AN. J Pediatr Orthop. 2012: doi:10.1097/BPO.0b013e318263a39e. 

Polly DW. Emerging technologies in intra-op guidance. Presented at: International Meeting on Advanced Spine Techniques; July 10-13, 2013; Vancouver, BC, Canada.

Sembrano JN. J Clin Neurosci. 2014; doi:10.1016/j.jocn.2013.04.011. Epub 2013 Oct 10.

Shin BJ. J Neurosurg Spine. 2012; doi:10.3171/2012.5.SPINE11399. Epub 2012 June 22. Review.

For more information:

Richard Fessler, MD, PhD, can be reached at Rush University Medical Center, Neurosurgery, Professional Building, 1725 W. Harrison St., Suite 855, Chicago, IL 60612; email:

A. Jay Khanna, MD, MBA, can be reached at Johns Hopkins Orthopaedic and Spine Surgery – National Capital Region, 6420 Rockledge Dr., Suite 2200, Bethesda, MD 20817; email:

Terrence T. Kim, MD, can be reached at The Spine Center, Cedars-Sinai Medical Center, 444 S. San Vicente Blvd., Mark Goodsen Bldg., Suite 800, Los Angeles, CA 90048; email:

David W. Polly, Jr., MD, can be reached at the Department of Orthopaedic Surgery, University of Minnesota, 2450 Riverside Ave. South, Suite R200, Minneapolis, MN 55454; email:

Charles A. Sansur, MD, can be reached at Department of Neurosurgery, University of Maryland School of Medicine, 22 S. Green St., S-12-D, Baltimore, MD 21201; email:

Jeff Siewerdsen, PhD, FAAPM, can be reached at Department of Biomedical Engineering, I-STAR Laboratory, Johns Hopkins University, Traylor Building, Room #718, 720 Rutland Ave., Baltimore, MD 21205; email:

Disclosures: Kim is a consultant for DePuy Synthes and has a research grant from Medtronic. Siewerdsen’s laboratory is supported by Siemens. Fessler, Khanna, Polly and Sansur have no relevant financial disclosures.



What is the current role, if any, for navigation in spine surgery?


Navigation will play role in safety, efficiency

Stephen M. Pirris, MD 

Stephen M. Pirris

Eric W. Nottmeier, MD 

Eric W. Nottmeier

Spinal navigation has an increasing role in improving the safety of complex spine surgery. At our center, we routinely use intraoperative spinal navigation and have an experience of more than 1,000 cases. Most of these surgeries involve spinal deformity and revision surgery. We have not had to return to the operating room to correct misplaced implants in more than 6 years. There are a multitude of peer-reviewed articles showing decreased pedicle breach rates when using navigation for screw placement.

Spinal navigation also increases the efficiency of surgery, and several studies have demonstrated a decrease in operating room time with the use of spinal image guidance as compared with traditional techniques. This technology can be utilized to help find bony structures through dense scar tissue as well as help to avoid critical neurovascular structures. The role of spinal navigation is expanding and can now be utilized in lateral thoracolumbar approaches to the spine and anterior approaches to revision or deformity cervical spine cases. Novel spinal fixation strategies can be employed that cannot be safely or precisely performed with traditional techniques. There are several recent publications describing these techniques.

The safety of intraoperative spinal navigation is not just applicable to patients, but also to surgeons and staff. Recently published data reveal the significant lack of radiation exposure to employees when utilizing spinal navigation. The use of spinal navigation also eliminates the need to wear heavy lead aprons. Wearing lead aprons can lead to degenerative joint disease at multiple sites in the body when frequently wearing them for a whole surgical career. 

Overall, the current role for spinal navigation is an important one and it will be increasing in the future. Patients will demand the safety value and surgeons will demand the safety and efficiency value.

Stephen M. Pirris, MD, is a consultant, assistant professor of neurosurgery and spine fellowship director at the Mayo Clinic, in Jacksonville, Fla.

Eric W. Nottmeier, MD,is a neurological surgeon at the Spine and Brain Institute of St. Vincent’s HealthCare, in Jacksonville, Fla.

Disclosures: Pirris has no relevant financial disclosures. Nottmeier receives speaking fees from Medtronic Navigation and Brainlab.


New technology should be used selectively, with caution

Mark M. Mikhael 

Mark M. Mikhael

Image-guided navigation techniques are being used in many spine surgery applications recently. Most of this technology is based on obtaining an intraoperative CT scan of the patient and using those images as a static representation of the patient’s spinal anatomy. This technology has been developed in hopes of increasing the accuracy of instrumentation placement, facilitating minimally invasive techniques through smaller incisions and assisting in complex deformity constructs. Although its utility in these instances has been encouraging, widespread and casual use of this innovative technology should be cautioned.

Although this technology decreases radiation exposure to the surgeon, assistants and operating room personnel, it does not spare the patient. More traditional techniques to place instrumentation using anatomical landmarks, tactile sensation and limited fluoroscopic imaging are often accurate and safe when used for short constructs in patients without significant deformity. While the decision to use this technology in long constructs may be reasonable when looking at the risk-benefit ratio to the patient, there may be an issue of overuse of this technology in more straightforward cases where the added exposure to the patient does not necessarily outweigh the benefits.

The use of image-guided navigation in minimally invasive spine surgery has also shown appeal because instrumentation can be accurately placed though small incisions without extensive exposure of the typical anatomy used for internal landmarks. Despite the level of sophistication of current imaging, limited exposures and specialized tools, it is imperative that the surgeon ensures the proposed goals of the surgical procedure are actually achieved. Placement of the instrumentation is only part of the task at hand. Proper bone preparation and graft placement are vital for successful outcomes and must not be sacrificed for the interest of percutaneous fixation.

Although the use of navigation can provide great assistance when performing long constructs, on complex deformities or upper cervical procedures, knowledge of traditional techniques and internal landmarks remains crucial. Oftentimes there is significant mobility throughout spinal segments, particularly in the upper cervical spine or in deformity surgery after releases/osteotomies, which renders the initial intraoperative CT-based image inaccurately calibrated to the instruments. To avoid serious injury to the patient, the surgeon must still have knowledge of the proper anatomic placement of instrumentation and cannot rely solely on the computer navigation display.

Enthusiasm over recent developments in navigation in spine surgery should be both welcomed and cautioned. As with all advances, we must continue to scrutinize our efforts to ensure patient safety and the efficacy of progress. It is also important to consider the changes that might make this technology viable for the future. Learned skills in this area will help the surgeon adapt to further progress in the field as outcome data prove the benefits and best indications in spine surgery.

Mark M. Mikhael, MD, is a reconstructive spine surgeon at Illinois Bone and Joint Institute and a clinician educator at the University of Chicago Pritzker School of Medicine.

Disclosure: Mikhael has no relevant financial disclosures.