Michael Ferris, a computer scientist with the University of Wisconsin-Madison, is working with medical physicists and oncologists at the University of Maryland Medical School on a computer program to reduce the threat of human error in setting radiation treatment plans.
Currently, doctors treating brain tumours with radiation surgery must develop a treatment plan that bombards the entire tumour, minimises exposure outside the target and avoids sensitive brain structures; a task that involves analysing scores of two-dimensional brain images within 40 minutes, as the patient waits in an uncomfortable head frame.
Ferris and his team aim to alleviate the pressure on doctors by optimising the Gamma Knife, a device that uses 201 radiation sources that combine simultaneously to create a ‘spherical ball’ of treatment.
With the Gamma Knife, each radiation shot is acts like a scalpel that burns out the tumour. When designing a treatment, doctors map out a series of radiation shots within the tumour, akin to filling a bag with marbles. But the spaces between those marbles end up eluding radiation exposure.
The computer-based optimisation by Ferris, however, takes a non-linear approach to the problem, said to better reflect how a radiation dosage truly behaves within a tumour.
Radiation attenuates or tapers out when it goes through material. The group recognised that the radiation is not 100 percent within the ball and zero outside of it, but is instead more of a bell-shaped curve with the highest dosage in the centre.
Using this bell-shaped design, the computer model was able to overlap different radiation shots to provide enough intensity to kill tumour cells in the spaces between doses.
The entire tumour needs to receive at least 50 percent of the maximum dosage to destroy the tumour. The computer program was also able to reduce the total number of doses needed for a completed treatment, which frequently can require ten or more trips through the Gamma Knife.
Another benefit of the computer program is speed. Since it is able to map out a treatment plan in 20 minutes or less, neurosurgeons can experiment with several alternative set-ups to find the most effective one.