PET/CT is not a larger gamma camera. What the positron scanner adds clinically, how far you can be from a cyclotron before the schedule stops being yours, and what the shielding, uptake rooms, staffing and quality assurance actually demand.
A hospital that has run a gamma camera for five years usually treats PET/CT as the next size up: a bigger machine, a bigger room, a bigger invoice. Very little carries across unchanged. The photon energy roughly triples, so the shielding calculation starts again from nothing. The tracer decays in under two hours instead of six, so the supply chain starts again. The patient becomes a walking source for an hour before the scan, so the floor plan starts again. What does transfer is the habit of keeping records, and that matters more than the hardware.
What follows is what has to be true before a PET/CT service works in a Bangladeshi hospital, in the order the decisions arrive. The choice between a plain gamma camera and SPECT/CT is a separate question, worked through in SPECT or SPECT/CT. This piece assumes you know why you own a gamma camera, and asks what a positron scanner adds on top of it.
A gamma camera throws most of its photons away. The collimator in front of the crystal accepts photons arriving from one direction and absorbs everything else, which is the price of knowing where a photon came from. PET does not need one. The two 511 keV photons produced when a positron annihilates leave in opposite directions, and the line joining the two detector elements that record them within a few nanoseconds of each other is the line the event came from. Removing the collimator is why PET sensitivity sits in a different class, and why a whole body acquisition takes tens of minutes rather than the best part of an hour.
Three things follow that a department notices in daily work.
What PET does not do is replace the SPECT list. Bone scintigraphy, renography, thyroid work, lung perfusion, sentinel node localisation and myocardial perfusion all stay on the gamma camera, and in most Bangladeshi departments they remain the majority of the workload. A PET/CT is an addition to a department, not a promotion, and a department that cannot keep a gamma camera in service will not keep a PET/CT in service either.
Fluorine 18 has a half-life of about 110 minutes. Everything awkward about running a PET service in a country with few cyclotrons follows from that number. Technetium 99m gives you a working day from one elution. FDG gives you a window, and the window is set by the distance between the cyclotron and your injection room.
You do not order an amount of FDG. You order an activity calibrated for a stated time, and the supplier's dispensing is worked backwards from it. If the first injection is at half past eight and the delivery arrives at half past ten, roughly half of what you paid for has gone before the first patient sees it. The useful question at planning stage is therefore not whether FDG is available in Bangladesh, but how long it takes to get from the cyclotron door to your hot lab, in Dhaka traffic, on a bad morning.
Work out the consequences on your own list before you commit to a scanner.
Gallium 68 changes this picture for neuroendocrine and prostate work, because the generator sits in your own hot lab and is eluted on demand. The trade is that generator yield falls across its life and it has to be replaced on a schedule, so cost per patient depends on how many patients go through each generator. Lutetium 177, which is where most departments planning PET are really heading, is dealt with in what a Lu-177 theranostics service needs, and ordering against a calibration time is set out in radiopharmaceutical supply and logistics.
The half value layer in lead at 140 keV is a fraction of a millimetre. At 511 keV it is several millimetres. A barrier sized for technetium is not a barrier for fluorine, and the mistake is not academic: departments have scaled their gamma camera drawings and found at survey that the wall into the reporting room fails. Three structural differences change the drawing.
On the bench side, the same energy shift is why PET rated hardware is heavier and made of tungsten rather than lead. Nuclear Shields supply a PET syringe shield range built around 8 mm of tungsten in the same barrel sizes as the technetium range, a Shielded Syringe Carrier PET with 10 mm of lead in the faces and 18 mm in the sides, and a Mobile PET Workstation with a 55 mm movable screen over a 20 mm shielded work surface. A Mobile PET Protection Barrier on castors, in 10, 20 and 30 mm, is what gives a technologist a shielded position beside the patient at injection. Room level shielding is covered in radiation shielding for a nuclear medicine department.
The most common planning error is treating the uptake period as waiting. It is part of the study. Muscle activity during uptake produces uptake in muscle, cold produces brown fat, talking produces vocal cord uptake, and a patient who walks to the toilet twice produces an image that is harder to report. An uptake room is a clinical space: shielded, warm, dimly lit, with a reclining chair, a call bell and a rule that the patient does not talk, chew or use a phone.
It is also what sets throughput, and the arithmetic is simple enough for the business case. Divide the uptake time by the time each patient occupies the scanner, round up, then add one position for the patient being injected. If uptake is an hour and the scanner slot is half an hour, three uptake positions are needed to keep one scanner working continuously. Build two and the scanner idles by design, whatever the sales projection says.
Two details are missed at drawing stage and cost money afterwards. The first is a dedicated toilet close to the uptake rooms, with its drainage arrangement agreed with the regulator rather than assumed: patients void immediately before the scan. The second is the route from the uptake room to the scanner. If it crosses the general waiting area, the shielding you calculated for the walls has been undone by the corridor.
A hot lab that works well for technetium needs three changes before it handles FDG: the bench shielding gets thicker, the handling gets faster, and the waste problem inverts.
Thicker is straightforward: 50 mm class workstations rather than the 20 or 30 mm that suits technetium, and tungsten instead of lead wherever the item is picked up. Faster matters because dose rate at the fingers during dispensing is the dominant occupational exposure in a PET department, and finger dose is driven almost entirely by how long the hands spend near an unshielded vial. Practised technique, a dose drawing station that holds the vial shield and the syringe in a fixed shielded position, and forceps rather than fingers, are what keep extremity readings sensible.
The waste inversion is the pleasant surprise. With a two hour half-life, decay in storage is close to trivial: what goes into the shielded bin on Sunday is background by Monday. What is not trivial is the first hour, when a swab or a syringe in an open bin is a live source at bench height. A shielded waste bin at the point of use is the whole answer. The layout logic for the room is in designing and equipping a radiopharmacy hot lab.
A PET service runs on a medical physicist who has done PET before, and on technologists trained on PET workflow rather than gamma camera workflow. Neither can be borrowed from radiology for an afternoon.
Recruit the physicist and at least two technologists early enough to be present through installation and acceptance testing. That period is the best training they will ever get and it costs nothing. The wider sequence is in staffing and training a nuclear medicine department.
PET quality assurance splits into work done before the first patient every day and work done on a longer cycle. The daily set is short: the manufacturer's own detector and coincidence check, the CT tube warm-up and air calibration, a dose calibrator constancy measurement and a background survey of the lab. None of it takes long. All of it is what stops a bad day becoming a bad quarter.
The periodic set is where the physicist earns the salary. Detector normalisation, CT number accuracy against a water phantom, PET to CT alignment, and the cross-calibration between the dose calibrator and the scanner. That last one decides whether a quantitative report is worth anything. If the dose calibrator reads five per cent high, every SUV the department issues is five per cent low, indefinitely, and nothing in the images will tell you. That is why the programme described in dose calibrator constancy, linearity and drift is a PET problem as much as a hot lab problem.
On the phantom side, Leeds Test Objects supply the PET IQ, an anthropomorphic PMMA body phantom with six hollow fillable spheres that screw into a plate on one face, filled with water for cold lesion imaging or with FDG for hot lesion imaging, and a set of PET CT MRI Cubes, seven clear PMMA cubes each holding a small fillable spherical void, used as point markers to check alignment between PET, CT and MR. Which phantoms to buy first, and in what order, is set out in choosing QA phantoms for a nuclear medicine department.
Everything that makes a gamma camera installation difficult in Bangladesh is worse for a PET/CT, because the CT sub-system draws a heavy transient load and the detector electronics dislike being switched off without warning.
The full site preparation sequence, which applies with more force here, is in siting a gamma camera.
A PET/CT that scans four patients a day is an expensive way to lose money, and the failure is rarely the machine. It is that the referral base was estimated from the size of the hospital rather than counted from the clinics. Before the business case is written, count for three months: how many patients would have had a PET, which clinician would have ordered it, and what they would have done differently with the result. If a large share of the projected list is staging that will not change management, the number is decoration.
Two other tests are worth applying honestly. If FDG has to travel from another city, and two deliveries a week arrive late enough to shorten the list, the service is not yours to schedule. And if the department is one technologist deep on the gamma camera, adding a PET/CT does not make it two deep. It makes the single point of failure more expensive.
Referring out is a legitimate long term arrangement rather than an admission of defeat, and it works when it is built instead of improvised: a named receiving centre, an agreed report turnaround, a route for images to come back in a form your physician can open, a clinician on your side who acts on the report, and a review of the numbers every six months. Departments that do this usually find that within two or three years the referral count either justifies a scanner or quietly demonstrates that it never will. Either answer is cheaper than finding out with the machine installed.
There is a middle path that is underused in Bangladesh. Quantitative SPECT/CT on a CZT system covers a useful part of what departments believe they need PET for, particularly absolute activity measurement for radionuclide therapy dosimetry, and it does it without a cyclotron anywhere in the supply chain. What that changes, configuration by configuration, is set out in which Spectrum Dynamics VERITON system fits your department.
And if a PET/CT is already installed and the real problem is that the processing is dated, replacing the console is a far smaller project than replacing the scanner. MiE supply the SCINTRON 7 PET/CT, a processing computer for installed Biograph PET/CT scanners that handles listmode acquisition, rebinning and reconstruction, applies prompt gamma, scatter and cascade corrections and runs online reconstruction so images can be reviewed before the patient leaves. There is also an ECAT SCINTRON upgrade for installed ECAT scanners, and ANCORIS, a dedicated cardiac positron scanner that acquires myocardial perfusion imaging and myocardial flow reserve in the same study.