Israeli team develops GPS accuracy for keyhole neurosurgery using mini-robot

Jerusalem 30 September 2007A tiny robot shimmies into position on the skull and reaches out its arm to automatically target the best spot for keyhole neurosurgery. It may sound like science fiction, or the latest animated feature by Dream Works, but it's actually an Israeli-developed mini-robot prototype that is likely to have a major impact on the way keyhole neurosurgery is done in the near future.


Keyhole neurosurgery, a minimally invasive procedure, is used for tumour biopsies, deep brain stimulation, to insert a draining tube to treat haematomas, and for catheter insertion. Increasingly, it will be used for tissue, tumour, and DNA sampling, which cannot be performed using anatomical imaging. Its advantage - that the surgery can be done with only a keyhole of 3-30 mm diameter opening in the skull - comes with a price. A slip in a surgical gesture is a surgery hazard.

"In neurosurgery, an error of two millimeters can be fatal or cause paralysis, loss of sight or memory, haemorrhages", stated Professor Leo Joskowicz of the Hebrew University, who along with two PhD students, Rudi Shamir and Moti Freiman, devised the sophisticated algorithms and programming that empowers MARS, a small, lightweight robot that has already revolutionized spinal surgery.

Developed by Professor Moshe Shoham of the Department of Mechanical Engineering, at the Technion in Haifa, the MARS, as programmed by Professorr Joskowicz, has already proven itself in delicate orthopaedic surgery via the SpineAssist, a device specifically for fusing vertebrae with screws in spinal surgery. The FDA-approved device, developed into a commercial product by Mazor Surgical Technologies, has already been used in over 400 spinal operations, and was ranked as the second "Clinical Achievement of 2005" by the prestigious Cleveland Clinic.

The move to take the MARS from orthopaedic surgery to keyhole surgery fell to Professor Joskowicz. Precise targeting of tumours, lesions, and anatomical structures cannot be done without support systems. Existing navigation support systems, using optical and mechanical guidance are limited by size, cost, and the requirement for manual passive arm positioning which can be time-consuming and error-prone. And the few robotic devices currently available are cumbersome and expensive.

"We had the platform; we needed to devise a new protocol according to the flow of surgery and the user's needs, and develop software", stated Professor Joskowicz, founder and head of the Computer-Aided Surgery and Medical Image Processing Laboratory at Hebrew University's School of Engineering and Computer Science.

The robot itself - encased in a sterile blue plastic wrap - weighs in at only 230 grams. For keyhole surgery of the skull, it was fitted with a rigid arm of 150 grams that can guide a needle, probe or catheter to the exact spot that the surgeon wants to target.

Targeting is the key word here, according to Professor Joskowicz. "Most neurosurgical gestures involve targeting. Every millimeter counts, because you work close to nerve roots. You can compare the novel neuro-surgical robot to a GPS. We invented a method to superimpose an MRI or CT image of the patient's brain over an image of the current surgical situation. But", he added, "it's not enough to have a map if you don't know where you are. The computerized robotic system connects the image to the surgical action, providing a quantitative link between the image and the surgical action", stated Professor Joskowicz.

How does it work? Although the MRI or CT image is usually done a day before surgery, it has to be aligned with an in-the-operating-room face scan - eyes, nose, ears. Automatic registration takes only a few seconds. Once the co-ordinates are aligned, the robot is positioned on the skull in the vicinity of the entry point. Using the combined image input, the robot is ready to move to the optimal position. "The surgeon sits at his PC and clicks on the image", stated Professor Joskowicz.

Professor Joskowicz brought out a "phantom" to demonstrate. A phantom, in professional lingo, is a plastic object that demonstrates a concept, in this case a plastic skull model with facial features - one of the students. An opening in the skull reveals "tumours". Affixing the miniature blue robot, Professor Joskowicz explained that an important task is to get the right trajectory, which is especially difficult in a small opening.

The robot positions itself and its arm, so that the arm is extended at just the right angle or trajectory to reach the target. The surgeon can then manually insert the needle, probe or catheter to the desired depth and have the Cranio-Assist make a small adjustment of the needle along the insertion axis.

"Our aim is to be able to use the little robot for many keyholes, minimal invasive surgeries (MIS) of the brain", stated Dr. Yigal Shoshan of the Department of Neurosurgery, School of Medicine, Hadassah University Hospital, who was part of the inter-disciplinary team that worked on the mini-robot algorithms. "It is fast, precise, and eliminates the problem of physiological tremors of a human being. As an MD, I think it is a very effective tool."

One of the benefits of the robotic system is that it is interactive. "If the surgeon wants to change the target and port of entry, all he has to do is click on the image", stated Dr. Shoshan. Using the co-ordinates, the operating system of the robot guides it into position for the keyhole surgery. "Assisted surgery still gives the doctor control", he added.

"We have tested the algorithms with humans in the operating room, without doing actual robotic-assisted surgery", stated Professor Joskowicz. "MARS Cranio-Assist is accurate. That is the key." Results of pre-operative CT/MRI and actual reality registration of the face and skull compare very favourably with other commercial navigated neurosurgery systems. Further, it ensures accuracy inside the brain.

Drilling a hole in the head, even a tiny hole, is a precision surgery task. The little robot not only increases surgical precision, but also causes less patient discomfort without a big frame clamped on the head. In some procedures the patient is given a local anaesthesia so that they can respond; easy with the non-obtrusive little robot.

"We hope to get the robot on the market within a year to 18 months", stated Ori Hadomi, CEO of Mazor Surgical Technologies. "No one has such a precise tool for keyhole brain surgery."

The innovative neurosurgery product made its United States debut at the Congress of Neurological Surgeons, held in San Diego in mid-September. In a private meeting organized by the medical giant Medtonic Inc., leading brain surgeons responded with great enthusiasm to a presentation of an updated prototype.

Professor Joskowicz, a native of Mexico, has been fascinated with robotics since he worked at IBM with Dr. Russell H. Taylor, who invented RoboDoc, a robot for total hip replacement developed in the United States. Professor Joskowicz founded the Hebrew University's Computer-Aided Surgery and Medical Image Processing Laboratory in 1996.

Professor Joskowicz' work on the mini-robot algorithms won him a Kaye Innovation Award, presented during the 70th Board of Governors meeting of Hebrew University last June. "I asked them to give the award to the team who worked on the development", stated Professor Joskowicz, modestly, referring to his PhD students, as well as Professor Shoham, Dr. Shoshan and Professor Felix Umansky of the Department of Neurosurgery at Hadassah Hebrew University Medical Center. More news on Professor Joskowicz' work is available in the VMW July 2007 article Miniature robot for precise positioning and targeting in keyhole neurosurgery wins Kaye Award for Hebrew University researcher.

Sharon Kanon

[Medical IT News][Calendar][Virtual Medical Worlds Community][News on Advanced IT]