The second detector does not make a better image, it makes a faster one. How to work the throughput arithmetic on your own study list, what dual-head ownership really costs, and when a single head is the correct buy.
The head count on a gamma camera is usually sold as an image quality question. It is not one. A single detector and a dual-detector system built on the same technology produce the same image from the same number of counts. What the second head buys is time, and time is what decides whether a department can absorb its own referral load without the list running into the evening.
That makes this a workload calculation, and it is one a hospital can do for itself before any vendor is in the room.
Notice what is missing from that list. Resolution, uniformity, energy resolution and count rate performance are properties of the detector, the collimator and the electronics. They do not improve because there are two of them.
The useful exercise takes an afternoon. Write down the studies you expect to run in a normal week, by type and number, at the volume you expect in year three rather than year one. Against each one, put a realistic table time for a single head and for a dual head. Then add the part everybody forgets: positioning, cannula check, patient on and off the table, collimator changes, and the reset between patients. That fixed overhead is identical on both machines, which is why doubling acquisition time does not double the working day.
Two things usually come out of that exercise. The first is that on a list dominated by planar and dynamic work, the single head loses very little. The second is that on a list with a lot of whole body bone scans and SPECT, it loses a great deal, and the loss is concentrated in the afternoon when the uptake period has finished and every injected patient is ready at once.
Throughput in nuclear medicine is often limited by the uptake area rather than the camera. Patients are injected, then wait. If you have one injection room and one waiting bay, a faster camera empties the bay quicker but the bottleneck moves upstream. Count your chairs as well as your detectors.
| Study | Does a second head help? |
|---|---|
| Whole body bone scan | Yes. Anterior and posterior in a single table pass rather than two. |
| SPECT of any region | Yes. Roughly half the acquisition time for the same counts. |
| Myocardial perfusion SPECT | Yes, but only if the gantry allows a 90 degree configuration. |
| Dynamic renal study | No. Acquired as a single posterior view; the second head is idle. |
| Thyroid uptake and scan | No. Single anterior view, often with a pinhole. |
| Static planar, single view | No. |
| Whole body iodine survey | Yes, for the same reason as the bone scan. |
| Lymphoscintigraphy for sentinel node | Marginal. Convenient for paired views, not decisive. |
Myocardial perfusion imaging is time-bound in a way that bone scanning is not. Stress and rest acquisitions have to happen in a defined relationship to each other and to the injection, the patient is in the department for hours, and a gated acquisition needs enough counts in every cardiac gate, so you cannot simply shorten the scan to catch up. On a single head, a cardiac SPECT acquisition takes roughly twice as long, and that lands squarely in the part of the day when the stress lab is producing patients.
If cardiology is the reason the hospital is buying, there are two honest answers and neither of them is a single-head general purpose camera. One is a dual head with a variable-angle gantry, which handles cardiac alongside general work. The other is a dedicated cardiac system built on digital CZT detectors, such as the Spectrum Dynamics D-SPECT from the USA, where the sensitivity gain changes the arithmetic altogether and lets the department trade the time saving for a lower administered activity instead. Setting up the surrounding service, the stress lab and the reporting pathway is a separate exercise, and a bigger one than choosing the camera.
For a district hospital or a first department building a service from nothing, a single head is often the correct engineering answer rather than a compromise. It is the right buy when the case-mix is dominated by thyroid uptake and imaging, static planar work, dynamic renal studies, hepatobiliary scans and paediatric renal cortical imaging, all of which are single-view acquisitions where a second detector contributes nothing. It is also the right buy when the projected volume simply does not fill a working day.
The secondary benefits are real: a smaller room, a lighter gantry and easier floor loading, one collimator set instead of two, one uniformity correction to maintain, and a simpler service contract.
The strongest argument, though, is what the difference in price buys instead. A properly designed hot lab with correct shielding, a voltage stabiliser and a UPS sized for the system, a set of QA phantoms, a service contract with a written response time, and applications training for two technologists rather than one. A dual head with no service cover and a single trained operator is a worse department than a single head with both, and it is a mistake this market makes regularly.
Most of the recurring cost of a gamma camera scales with detector count, and the items are not always visible in the quotation.
Ask specifically how collimators are exchanged. On a mixed list a department may change collimators several times a day, and whether that is a two-person manual lift or an assisted exchange on a cart makes a measurable difference to both the schedule and the injury risk.
If the answer to the throughput question is that a single head fits your projected week with an hour of slack in it, buy the single head and spend the difference on the parts of the department that cannot be retrofitted.