The easiest way to look inside the body is to do a quick CT scan. If you need something better, you can make a reservation for an injection of liquid metal contrast and a cage-rattling MRI. But if you need something more than a glimpse of the anatomy, an actual comparative look at which tissues are under-performing, and which are over-performing (as in cancer), you will probably need a PET scan.
Generating a PET (Positron Emission Tomography) image requires an injection with a modified version of some essential molecule in the particular metabolic pathway you want to examine. To see rapidly metabolizing tumors that might be taking up a lot of glucose, for example, you get a radioactive version of that sugar known as fluorodeoxyglucose. When this tracer is trapped in presumptive cancer cells, its positrons eventually decay and they fire off two gamma rays in exactly opposite directions. Scintillating crystals downshift the energy from those gammas into the optical range, which can then be converted into an electrical signal using an appropriate photodetector.
The secret to getting pixels or voxels out of all this is to gate your detectors for precise coincidence between the two opposed gammas. This requires speed — 300-picosecond speed. This timing accuracy then allows for the localization in 3D space of the point of origin for the initial decay.
The fundamental problem with current PET machines is that you are given a relatively massive tracer dose, but then you are scanned with a relatively narrow detector ring that slowly plods along collecting the gammas in mere 20-cm-width segments. For a large bore scanner, that’s a relatively small detector cross section compared with the full 360-degree spherical emission range.
What this means is that your body is emitting precious radioactive signal (precious in that you are paying for it with the initial absorbed dose), but the underpowered hardware is letting the bulk of it slip out the sides and go to waste. Stated another way, it is the patient that ultimately pays the price of a poorly matched tracer and detector — both in terms of how long they need to remained stuffed inside the machine, and how many times they can safely be subjected to the ordeal.
The solution to this problem, one now embraced by a collaborative effort funded by the DOE to the tune of $15.5 million, is to scan the whole body at once. Recognizing the aforementioned truths, several cooler heads have gotten together to provide for any and all that which industry alone has so far failed to deliver. The project sprang from an effort at Berkeley originally called OpenPET, which had the noble aim of opening and democratizing PET electronics. High-speed coincidence detectors are nothing new; they are the bread-and-butter of all sorts of high-energy physics mega-projects. For that matter, so is the scintillator and photodetector hardware. OpenPET has now evolved into the Explorer project, which will produce a PET machine with an unprecedented half-a-million detectors.
Via Gust MEES, Dr. Stefan Gruenwald