Year: 2007

Independent artificial intelligence researcher Jeff Hawkins has an article in this month’s IEEE Spectrum magazine asking the question ‘why can’t a computer be more like a brain?’.

Hawkins argues that while we hope that machines will be able to simulate human intelligence, we ignore the thing that makes us so – the brain.

He suggests that we need to create artificial intelligence systems that closely match the architecture of the brain to achieve this task.

Hawkins has outlined his arguments, and his own theories of simulated brain architecture, in his book On Intelligence, but if you want a whistle-stop tour of his theories, this article is a great summary.

Link to Hawkin’s article ‘Learn Like a Human’.

Researchers from the University of Calgary have released the first version of NeuroArm – a surgeon-controlled robot for conducting brain surgery.

The key feature of the robot is that it is designed to work inside an MRI brain scanner.

MRI scans currently provide the most accurate structural image of the brain and therefore provide important information for planning operations.

Neurosurgeons also use MRI scans completed before surgery to guide the operation while it’s happening, using a method called stereotactic neurosurgery.

This allows surgical instruments to be guided to an exact spot in the brain by tracking their position in real time, in relation to the 3D scan completed earlier.

One disadvantage is that the brain scan can’t be updated as the brain is altered during the operation.

Being able to scan people while they’re having surgery might sound a simple idea, but MRI scans involve the patient being inside a tube surrounded by hugely powerful magnets, meaning the environment isn’t accommodating to surgeons who need free space and surgical steel.

NeuroArm has been designed to fit inside the tube, and crucially, is not made of any magnetic materials that will affect and be affected by the MRI machine.

This means the surgeon can update the brain scan and complete the operation by controlling the robot remotely.

He or she can do this by using a specially designed surgical workstation that provides a virtual interface to the robot arms, including force feedback on the tools, so the surgeon does not have to give up his ‘surgical touch’.

While the current set-up seems to involve the surgeon being located in the same building as the patient, it is interesting to speculate that, in the future, operations could be directed from hundreds or even thousands of miles away.

The combination of the accurate brain scan and the robot controlled tools also means that the surgeon should be able to attempt microsurgery on very fine brain structures.

You may be surprised to learn that robot-assisted neurosurgery isn’t particularly new and was introduced in the 1980s.

Brown University has a fantastic history of the technology and procedures if you want some background.

Link to Project neuroArm page.
Link to more info on Project neuroArm.
Link to write-up from New Scientist.
Link to history of robotic neurosurgery page.

In 1955, after seven years of trying, John and Mary’s first child was born. The birth of Casey Holter turned John Holter’s life upside-down and changed the course of medical history.

Agonisingly, Casey had spina bifida, a condition where the spine doesn’t fully form and may be dangerously misshapen.

The condition was also causing hydrocephalus, a life-threatening build-up of fluid in the brain.

The fluid that surrounds the brain is called cerebrospinal fluid or CSF and acts as a fluid ‘bath’ which cushions and protects the delicate organ.

It is produced by a structure in the brain stem called the choroid plexus and circulates around the brain before being drained into the blood supply.

If the drainage system is blocked, however, it can lead to a dangerous build-up that can pressure, distort and eventually damage the brain beyond repair. If left untreated, it can be deadly.

In 1955, the only thing keeping Casey Holter alive was a twice daily procedure where a needle was inserted into the fontanelle, the soft spot on a baby’s head, and the excess fluid was removed with a syringe to reduce the pressure.

Eventually, Casey was given an operation by the neurosurgeon Eugene Spitz to insert a ball and spring valve that would, in principle, allow the fluid to drain into the blood supply, without letting anything dangerous from the blood wash back into the CSF.

Unfortunately, the valve was clumsy technology, and when inserted, it irritated Casey’s heart to the point where the young child had a heart attack and suffered permanent brain damage.

John Holter, then working as a technician in a hydraulics factory, asked Eugene Spitz about the details of the procedure. He was surprised that the problem, which seemed to him like a simple hydraulics issue, had not been solved.

He had noticed that when nurses inserted needles into certain types of medical tubing, leaks didn’t occur because the gap was water-tight under low pressure conditions.

But, like a teat on a baby’s bottle, when the pressure was high enough the gap opened and the fluid forced its way through. A perfect valve for releasing built-up CSF and preventing backwash.

Holter went home, sat in his workshop, and constructed the first version that very evening. It was a rough-and-ready rubber-tubing and condom prototype, but it worked.

While the principle was sound, Spitz noted that that the valve must made of a material that wouldn’t irritate the body, as this might cause the same problem that had brain-damaged his son.

Holter contacted Dow Chemical and was advised to use silicone, at the time, a newly developed material.

Holter had created a usable version within months. So quickly, in fact, that his son was still too weak from the last operation to have it installed.

It was first and successfully installed in another child, and then in March 1956, Eugene Spitz installed John Holter’s valve into Casey, successfully treating his hydrocephalus.

Sadly, Casey never fully recovered from his brain damage from the initial operation, and died during an epileptic seizure five years later.

Fittingly, Casey’s legacy is that Holter’s invention, now called the Spitz-Holter shunt, is still in use today.

Holter spent the rest of his life developing valves for medical use and passed away in 2003, having saved the lives of thousands children affected by the same condition as his son.

It is estimated that 15,000 valves based on Holter’s design are installed every year in the United States alone.

John Holter’s remarkable story was retold in a 2001 paper published in the Journal of the American College of Surgeons upon which this article was based.

Link to PubMed entry for Journal of the American College of Surgeons paper.