Some tumors are invisible on an X-ray scan because they look just like the tissue around them. Tumor cells can be made to glow faintly, the way a firefly does, a property called bioluminescence. This device sees that glow from several sides and works out where the tumor is, so the machine can aim at it.
I led turning a Johns Hopkins research prototype into a desktop-computer-sized product, and we brought it to market in 2015.
Tumor cells that make their own light
Light from inside tissue spreads out before it reaches the surface, so a single picture shows only a blur. The bed slides into a dark imaging chamber, where a set of mirrors turns 360° around it and carries its glow, from whichever side it faces, to a fixed mirror, a filter wheel and a single cooled camera. Matched to the outline from the CT scan, software works back to where the light started, and that point becomes the target.
From a lab prototype to a desktop product
It grew out of a National Institutes of Health partnership (2011–2015) between Johns Hopkins and us. The prototype was an optical assembly under a light-tight dome that had to be docked onto the irradiator by hand, built around a large research camera. We made it a desktop-sized, light-tight unit that works on its own on a bench, with a removable bed that carries the subject straight to the irradiator, a fourth mirror to keep it compact, a new camera (the Andor iKon-M 934) and CE marking.