DoseOptics

Here is a fact about radiation therapy that is either obvious or unsettling depending on how long you think about it: for most of its history, nobody could see it. A machine points a beam of high-energy photons at a tumor, the beam does its work, the patient goes home, and comes back tomorrow to do it again, twenty or thirty times. The dose is invisible. The team delivering it is, in a real sense, aiming in the dark and trusting the plan. DoseOptics is a company built on the premise that this is a strange way to run one of the most common cancer treatments on earth, and that there is, in fact, a way to turn the lights on.

The way involves a phenomenon with a wonderful name: Cherenkov light. When a charged particle moves through tissue faster than light moves through that same tissue, it emits a faint glow - the same eerie blue you have seen in photographs of nuclear reactor pools. It turns out that a radiotherapy beam passing through a patient produces this glow too, just extraordinarily faintly. DoseOptics builds cameras sensitive enough to catch it. Point one at a patient during treatment and you get something that did not exist before: a live video of the dose, landing where it lands, as it happens.

Every good deep-tech company has an origin story that sounds too neat to be true, and DoseOptics has two. The first is a hallway moment. Brian Pogue, a biomedical engineering professor at Dartmouth's Thayer School of Engineering, was at a conference when he heard a colleague present on Cherenkov radiation and imaging. By several accounts he understood almost immediately that the same physics could be pointed at radiation therapy. The insight was not that Cherenkov light exists - physicists have known that for the better part of a century - but that it could be turned into a clinical instrument.

The second is a motivation. In 2010, The New York Times published a series of articles documenting cases where radiation therapy had gone wrong - dosing errors, misdirected beams, patients harmed by a treatment that was supposed to help them. The reporting landed on a lot of desks. On Pogue's, it seems to have become a design brief. If the problem was that nobody could see what the beam was doing, then the fix, at least conceptually, was to build something that could.

The technology was developed at Dartmouth, financed in its early years largely by grants from the National Cancer Institute at the NIH, and then licensed out to the company Pogue co-founded in 2015 with William Ware, who serves as CEO. This is the academic-to-clinic pipeline working exactly as the brochures promise: invention at a university, a spinout to commercialize it, and eventually a product installed in treatment rooms. It does not always go this smoothly. Here it more or less did.

What is worth pausing on is the funding path. DoseOptics raised some founder seed money and then, notably, went to the government rather than to venture capital. The bulk of its early money came from NIH grants, including a $2 million award in 2016 for further clinical trials, bringing the total to roughly $3.4 million. That is an unusual capital structure for a medical-device startup, and it tells you something about the company - patient, research-driven, comfortable operating on grant timelines rather than growth-round urgency. Deep tech aimed at safety does not always fit the venture template, and DoseOptics did not try to force it.