The whole lifecycle in the order it has to happen: the case-mix that justifies the service, the licence, the room, the camera, the hot lab, the QA programme, the people, and what keeps it running afterwards.
The board approves the camera first. That is the usual order of events in a Bangladeshi private hospital, and it is why so many nuclear medicine departments end up altering a room that was finished six months earlier: a hot lab squeezed into a corner because it was not on the original drawing, a delay tank nobody costed, injected patients waiting on the same bench as outpatients because there was only ever one waiting area.
Nuclear medicine is unusual among hospital services in that the building, the licence, the equipment, the pharmacy and the people have to arrive in a particular sequence. Get one of them out of order and the cost is not inconvenience, it is rework on a shielded room or a camera sitting in a crate waiting for a permission. That sequence runs from the referral base that justifies the service through to the service contract that keeps it alive in year five, and each stage of it has been treated in depth on its own elsewhere in this section. What matters at this level is how they connect.
The honest first exercise is not a market study. It is counting the patients your own consultants currently send somewhere else. Ask oncology how many bone scans they refer out in a month, urology and nephrology how many renograms and DMSA studies, endocrinology how many thyroid uptakes and how many patients they send for radioiodine, cardiology how many myocardial perfusion studies, and general surgery how many sentinel node cases they would do if the tracer were available in-house.
That count matters more than the total number of beds, because a gamma camera is a workload machine. Its economics are set by how many studies pass through it per working day, and its clinical value is set by whether the referrers trust the report. A department with a strong oncology and orthopaedic base will fill a camera on bone scintigraphy alone. A department whose case-mix is mostly cardiac has a different equipment answer entirely. A department with neither, buying because a competitor bought, will run three studies a day and blame the machine.
Case-mix also decides how much building you need. Diagnostic imaging alone is one set of requirements. Adding radioiodine therapy for thyroid disease adds inpatient rooms with their own shielding, their own drainage and their own discharge rules, and that decision has to be made at drawing stage rather than added later. Paediatric work adds sedation and immobilisation requirements. Sentinel node work adds a coordination burden with theatre. None of these is a reason not to proceed, but each of them changes the brief you give the architect.
A nuclear medicine facility in Bangladesh operates under authorisation from the national nuclear regulator, and the practical consequence for a project manager is that several approvals sit upstream of construction, not downstream of it. The shielding design is reviewed before the walls go up. The authorisation to import and possess radioactive material is separate from the permission to operate the facility, and both are separate from the customs clearance of the equipment itself.
Three things are worth doing early, because they are the ones that hold projects up. Appoint the radiation protection officer at the start rather than the week before the licence application, because that person has to be trained and named and will be the signature on much of the paperwork. Get the medical physicist involved at drawing stage, because the shielding calculation is their document and the architect cannot produce it. And ask the regulator directly for the current forms and current expectations rather than relying on what a colleague did three years ago, because requirements are revised and a rejected application costs a month.
Radiation safety and regulatory compliance for a Bangladeshi nuclear medicine department is a subject in its own right, covering dosimetry, source records, waste, transport and inspection. Treat it as a parallel workstream with its own owner from the day the project starts, rather than as a document produced at the end.
A gamma camera gantry is a concentrated load, and the structural engineer needs the manufacturer's load plan, not a general figure. Ground floor is easier for that reason and for the delivery route. Before anything is ordered, walk the route the crate will take from the hospital gate to the room and measure it: gate width, ramp, corridor turns, door openings, lift car dimensions and lift capacity if the room is not on the ground floor. This is the most common and most avoidable delay on a Bangladeshi installation, and it is discovered by tape measure, not by drawing.
Shielding is decided before the concrete is poured. Lead or barite thickness depends on the isotopes in use, the activities handled, the occupancy of the spaces on the other side of each wall and, importantly, the spaces above and below. An office directly over a hot lab is a shielding problem that costs far more to solve after the slab is cast. The hot lab, the injection room, the uptake or waiting room for injected patients and the imaging room want to be adjacent to each other and away from general circulation, so that a dispensed dose travels the shortest possible distance in the open.
After the structure comes the services. Dedicated air conditioning for the equipment, sized to run continuously and not shared with a corridor. Humidity control, because monsoon humidity is hard on detector electronics and on any equipment left off overnight in an unconditioned room. Clean, stable power with a voltage stabiliser and a correctly sized UPS, because Bangladeshi supply is not gentle and repeated hard shutdowns during a gantry movement are how boards die. Drainage for a delay tank if therapy is planned. Ceiling height and clear ceiling structure if room lasers are to be fitted for consistent patient positioning, which is worth doing at build stage rather than retrofitting into a finished ceiling.
The equipment decision follows the case-mix, and it is really four decisions taken together.
Systems from MiE GmbH of Germany, including the SCINTRON range with its acquisition and processing modules, and the digital CZT platforms from Spectrum Dynamics Medical of the USA, cover different points on that grid. The mistake to avoid is specifying a system around one flagship study while the department's actual daily list is bone scans and renograms.
Whatever is chosen, the workstation and processing software deserve a real specification rather than a single line at the bottom of the tender. Reconstruction options, the clinical module set, DICOM behaviour with your PACS and reporting are what the technologist and the reporting physician touch every day.
The hot lab is small, and its cost per square metre is high. It carries the generator, the dispensing station with an L-block, lead bench shielding, a fume hood or laminar flow cabinet depending on what is being prepared, the dose calibrator, syringe and vial shields, a contamination monitor, a shielded waste store for decay-in-storage, and the sink and decontamination provision that everyone forgets until the first spill.
Buy the shielding as a designed set rather than assembling it from whatever is available. Shielding equipment from Nuclear Shields of the Netherlands is specified as a workflow: where the vial sits, where the syringe is filled, where the assay happens, where the dispensed dose is carried from. A dispensing bench that forces the technologist to reach around an L-block is a daily dose problem, and it will show on the finger dosimeters.
Two operational points follow from the hot lab and are worth deciding at design stage. First, receipt and dispensing should be separable, so that a delivery arriving mid-list does not interrupt a dispensing operation. Second, the waste store needs to be sized for the longest-lived isotope you will use, held for enough half-lives, with the records to prove it. Undersizing that room is the classic error, and the fix is expensive because it is a shielded room.
Acceptance testing and commissioning establish that the system meets what was bought. The QA programme establishes that it still does. They are different activities and the second one is the one that gets neglected, usually until an inspection or until images start looking wrong and there is no baseline to compare against.
The equipment for it is modest and should be ordered with the camera, not afterwards. A flood source and flood phantom for uniformity, a bar pattern phantom for spatial resolution and linearity, a hole or grid plate for distortion, and a SPECT performance phantom for reconstructed image quality, contrast and uniformity. Leeds Test Objects of the UK make phantoms across gamma camera, SPECT, PET, CT and general radiology, which matters if the department is also going to be responsible for the CT half of a SPECT/CT.
The programme itself is a schedule, an owner and a record: daily uniformity and energy peak, weekly resolution and linearity, periodic centre of rotation and SPECT phantom, and the dose calibrator's own constancy, accuracy, linearity and geometry checks. What makes it survive an inspection is not the tests, it is the fact that the results were plotted over time and that somebody acted on a trend. Choosing which phantoms to buy first, and building a QA programme that holds up, are covered separately and in more detail.
A working department needs a physician trained in nuclear medicine to report, a medical physicist for dosimetry, shielding, QA oversight and the regulatory file, technologists to acquire and process, someone competent in the hot lab, a nurse for cannulation and patient care, and a named radiation protection officer, which is a role rather than a separate post. In a small department several of these are the same two or three people.
That is where the risk sits. A department where one technologist knows the acquisition protocols and one person can dispense will lose its list the first time either of them takes leave. Cross-train from the beginning, and write the applications training in the purchase contract for at least two people rather than one, with a refresher visit some months after go-live when the staff have real questions. Planning the first year of staffing and training is worth doing formally, because recruitment in this specialty takes longer than equipment delivery.
Commissioning is a sequence, and the order protects you.
Do not schedule the first patient before the first radiopharmaceutical delivery has been arranged and tested. A camera that is accepted and licensed and has no technetium is not a working department, and the supply chain is the piece most often left until last.
After the opening ceremony the department is judged on uptime and on turnaround. Both come down to a small number of unglamorous things.
The departments that work well in Bangladesh are not always the ones with the newest camera. They are the ones where the hot lab was designed rather than fitted in, where two people can do every job, where the QA book is current, and where somebody owns the isotope order the way a pharmacy owns its stock.