How Gamma Knife Works: Cobalt-60 Beams Converging on a Target, in Plain Terms

Gamma Knife works by aiming about 192 individually weak beams of cobalt-60 gamma radiation so they all cross at one point in the brain, where their combined dose is high enough to treat the target while the tissue each beam passes through is barely touched. Each single beam is too weak to do harm on its own. Only at the crossing point, called the isocentre, do the doses add up. That is the whole trick, and once it clicked for me it stopped feeling like magic and started feeling like arithmetic¹.

When my acoustic neuroma was found, “focused radiation” was the phrase everyone used, and it told me almost nothing. What I wanted to know was how something with no moving blade could reach a lump behind my ear without damaging everything in front of it. This is the plain-terms answer I eventually pieced together, checked against the physics rather than the brochures. For the wider picture of the treatment and what it treats, start with Gamma Knife radiosurgery; for the category it belongs to, see what is stereotactic radiosurgery.

How does Gamma Knife actually work?

Gamma Knife works on a principle of convergence: many weak beams enter the head from different directions and meet at one target point, so the dose is trivial along each beam path but high where they all cross. Picture sunlight through a magnifying glass. The light falling on your hand is harmless, but focus it to a point and the paper there catches. Gamma Knife does the same with gamma rays, except it uses about 192 beams from fixed positions all around the head rather than one lens².

Because the beams come from many angles, no single line through healthy brain ever carries more than a small fraction of the treatment dose. The target, sitting where all 192 lines intersect, receives the full effect. This is why it can treat lesions roughly 5 to 40 mm across while leaving the surrounding brain largely alone¹. It is not surgery: there is no incision and no knife, and I was awake and talking the entire time.

What are the cobalt-60 sources and how many beams are there?

The radiation comes from cobalt-60, a radioactive form of cobalt that gives off gamma rays; current Gamma Knife models use 192 of these sources, while older units used 201. That is why you will see both numbers in different places. Both refer to the same idea: a fixed array of small sealed sources arranged around the head, each sending one weak beam inward toward the shared target³.

Unlike a machine that swings an arm around you, the Gamma Knife sources do not move during a shot. They are built into the unit in a fixed geometry, which is part of why the convergence is so reliable. The Leksell Gamma Knife is made by Elekta and traces back to the neurosurgeon Lars Leksell, who proposed the concept in 1968; current models include Perfexion, Leksell Icon and Gamma Knife Esprit³. The cobalt is sealed and stays in the machine. Nothing radioactive goes into you, and you are not radioactive afterwards.

What is the isocentre and how precise is it?

The isocentre is the single point where all the beams cross, and Gamma Knife focuses them there to an accuracy of under about 0.5 mm. That sub-millimetre precision is the reason a target can sit right beside a nerve, a blood vessel or the brainstem and still be treated, because the dose falls away steeply just outside the crossing point¹.

For me this was the reassuring part. My tumour was close to the facial and hearing nerves, and the whole point of the technique is that the sharp edge of the dose can be placed between the target and the structures you want to protect. A simple, roughly round target may need only one isocentre. The precision is also why so much of the day is spent on imaging and planning rather than on an operation, which is covered in Gamma Knife planning and dose. The frame or mask that holds your head still enough to hit that half-millimetre is explained in frame-based versus frameless radiosurgery.

What do collimators do?

Collimators are the openings that shape each beam and set how wide the treated point is, so the team can make the high-dose zone small or large and control how sharply the dose falls off. A Gamma Knife typically offers a few collimator sizes, for example 4, 8 and 16 millimetres, and the plan chooses which to use for each shot².

Think of them as different-sized nozzles. A small collimator makes a tight, sharply bounded point of dose, useful for a tiny target or for tucking the edge in close to something delicate. A larger collimator covers more ground in one shot. In practice the team mixes sizes, and modern units can even block or reshape individual beams to steer the dose away from a critical structure. All of this happens in the planning stage while you wait, part of the “long middle” of the day that nobody warned me about, described in the day of Gamma Knife, hour by hour.

How do multiple isocentres shape an awkward target?

Real tumours are rarely neat spheres, so the plan builds the treated region out of several isocentres of different sizes, packed together like beads until the combined high-dose zone matches the shape of the target. One isocentre makes a roughly round pocket of dose; a handful, carefully placed and sized, can be blended into an irregular shape that hugs a lobed or elongated lesion⁴.

This is where the medical physicist earns their place on the team. Each isocentre is positioned and weighted so the edges overlap correctly and the dose stays high across the whole target while dropping off fast just beyond it. When the physicist showed me my plan on screen, it looked less like a beam of light and more like a cluster of overlapping bubbles filling in the outline of my tumour, and that image finally made the technique concrete for me. The team that does this, a neurosurgeon, a radiation oncologist and a physicist, and the dose they aim for are set out in Gamma Knife planning and dose.

Why does the physics mean it works slowly?

Gamma Knife delivers its dose in minutes, but the biological effect unfolds over months to years, because the radiation damages the target’s cells so they stop dividing rather than removing the lesion on the spot. The target is usually still visible on the first follow-up scans. Control, meaning it stops growing, not disappearance, is the goal⁵.

This surprised me more than anything. I had imagined walking out treated and finished, but benign tumours change slowly over about 1 to 3 years, arteriovenous malformations close over 2 to 3 years, and trigeminal pain eases over days to about 2 months. A single high dose in one sitting is “radiosurgery”; when the same idea is split into 2 to 5 sessions it is called “stereotactic radiotherapy”⁵. Because nothing dramatic happens on the day, the waiting between scans afterwards was, for me, the hardest part; I write about that in radiosurgery and scanxiety, and what the results and follow-up actually look like in Gamma Knife results and follow-up.