Nuclear medicine teleradiology: authorized user rules

Nuclear medicine is where imaging center coverage plans go thin. The volume is modest, the reader supply is the scarcest in radiology, and the regulatory vocabulary, authorized users, radioactive materials licenses, Agreement States, makes buyers hesitate to outsource reads that are otherwise well suited to remote interpretation.

Nuclear medicine teleradiology works, and facilities run it every day. What separates a clean arrangement from a messy one is knowing exactly which responsibilities move to the remote reader and which are pinned to your license by federal regulation. This page draws that line first, then covers the study types, the workflow differences from CT and MRI teleradiology, and the vendor questions specific to this modality. For a coverage plan against your own nuclear volume, a written rate card follows a request within one business day.

Where do the NRC authorized user rules actually apply?

Start with what the rules govern. The Nuclear Regulatory Commission regulates the medical use of byproduct material under 10 CFR Part 35: the possession of radiopharmaceuticals, their ordering and administration, and the people permitted to direct that use. An authorized user is a physician who meets the training requirements of Part 35 and is named for that use on a radioactive materials license. For imaging and localization studies, 10 CFR 35.290 sets the training bar: certification by a specialty board recognized by the NRC or an Agreement State, or 700 documented hours of training and experience including at least 80 hours of classroom and laboratory work in radionuclide handling. Enforcement runs through the NRC and the 40 Agreement States that regulate materials licensees within their borders under compatible rules.

Now the line that matters for teleradiology: those rules attach to the use of the material, and the use happens at your facility. The injection of the tracer, the supervision of that administration, the written directives where required, the radiation safety program behind all of it, every piece stays with the authorized user on your license and your radiation safety officer, no matter who reads the images. Interpretation of the transmitted study is the practice of medicine, governed by state medical licensure, and it is the one component a teleradiology arrangement moves.

That distinction cuts both ways, and both matter in procurement. It means a remote radiologist does not need to appear on your radioactive materials license to interpret your bone scans, so the licensing objection that sometimes kills these arrangements early is not the regulation speaking. It equally means outsourcing reads outsources none of your Part 35 obligations, so a vendor pitch that gestures at handling "nuclear compliance" is describing something the regulatory structure does not let it hold. Your bylaws, your accreditation program, and your payers will still expect documented nuclear training in whoever interprets, which is a credentialing question answered with board certificates and fellowship files, and none of it involves your license paperwork. Confirm the division with your radiation safety officer and counsel; this page is not legal advice, and the regulatory text is current as of September 2026, with an NRC proposed modernization of the byproduct material rules published in May 2026 and worth watching.

Where authorized user status does bear on a reading service is reader quality. The board pathway in 35.290 is the route most nuclear radiologists and nuclear medicine physicians take, so AU-eligibility is a compact proxy for real nuclear training. AstraRad treats nuclear medicine and PET-CT as one of its ten subspecialties, staffed by board-certified, fellowship-trained readers, and routes every gamma camera and PET-CT study there by default; the modality service is described on PET-CT and nuclear medicine reads.

Which nuclear medicine studies route remotely well?

Nearly all diagnostic nuclear medicine transmits and reads remotely, because the interpretation depends on the acquired data set; presence at the camera adds nothing once the acquisition is done. The useful sorting is by what has to travel with the images and by where the read usually lives.

Study type	Routes remotely	What the remote reader needs with the images
Bone scintigraphy, whole body and SPECT	Well	Priors above all; flow and delayed phases labeled
Thyroid uptake and scan	Well	Uptake values and probe measurements, plus TSH and medication history
Renal scintigraphy	Well, with the data	Dynamic series plus processed curves, split function, and clearance values
Hepatobiliary studies	Well, with the data	Full dynamic set, timing of morphine or CCK if given, ejection fraction
Gastric emptying	Well, with the data	Standardized meal documentation and the retention percentages by hour
Lung ventilation-perfusion	Well, often urgent	Recent chest X-ray for comparison and the clinical probability context
Cardiac SPECT perfusion	Variable	Processed slices and quantitative output; at many facilities this read sits with cardiology
PET-CT, oncologic	Well	Priors, pathology and treatment history, and a reader trained on both components

Two rows carry most of the friction. Cardiac SPECT perfusion is a coverage-model question before a technical one: at many facilities the myocardial perfusion volume is read by cardiology under its own arrangements, and a teleradiology contract should state plainly whether that stream is in or out of scope. PET-CT is the opposite: firmly radiology-side work, growing volume, and the study type where subspecialty depth shows most, because an oncologic staging read synthesizes tracer distribution and anatomy into one narrative a tumor board can act on. How matched routing works in general is covered in subspecialty teleradiology services, and the credential-verification method in teleradiology specialists.

Turnaround expectations should follow the clinical tempo rather than a uniform SLA reflex. Most of this modality is scheduled work read comfortably at routine tier. The urgent exceptions are real: a ventilation-perfusion study for suspected pulmonary embolism in a patient who cannot receive CT contrast is a same-hour question, and hepatobiliary and gastrointestinal bleeding studies can be. A clean contract prices nuclear medicine at routine by default with a printed priority multiplier for escalation, so the exceptional case is a workflow choice instead of a negotiation.

Why the nuclear bench is the thinnest in radiology, and what that does to your options

The staffing math behind this modality explains most of what buyers experience in it. A typical imaging center's nuclear volume is a fraction of its CT volume, which makes the coverage need real but part-time, and part-time subspecialty needs are the hardest to hire for. The reader pool is small to begin with: nuclear medicine and nuclear radiology together produce few new physicians a year relative to the large fellowships, and the 700-hour or board-pathway training bar in 35.290 keeps the supply tight by design. A facility that loses its one nuclear-trained reader is often looking at months of recruiting for a role that was never a full FTE of work.

That scarcity shapes the outsourcing decision in two ways. It makes remote reading the structurally sensible answer for most sub-hospital nuclear volumes, because a teleradiology panel amortizes one scarce reader across many facilities' part-time needs. And it makes panel depth the first thing to probe in any vendor conversation, because the same scarcity that emptied your bench constrains theirs. Two readers carrying a vendor's entire nuclear stream is a common and fragile arrangement; ask for the count, the trailing-year nuclear case volume per reader, and what happens to your PET-CT turnaround when one of them takes leave. The generalization of that fragility argument, and the regional supply picture behind it, is in the radiologist shortage by geography.

How does nuclear medicine teleradiology differ from CT and MRI workflow?

The transport layer is the same DICOM push a CT takes. Everything else about the hand-off is heavier, and underestimating that is the most common implementation failure in this modality.

The processed data is part of the examination. A renal study without its curves, a gastric emptying study without its retention percentages, a cardiac study without its quantitative output is half an examination, and a remote reader handed only raw dynamic frames will either re-process, hedge, or call your technologist at 4 p.m. Configure the camera workstations to export processed series and secondary captures with every study, and test the export path modality by modality during onboarding, before the gaps can surface case by case in live reads.

The tracer metadata must travel too. A nuclear report states the radiopharmaceutical, the administered activity, and the route; a reader without the dose record cannot write a complete report. This is order-entry and RIS configuration work on your side, done once.

Protocols span days and interruptions. Multi-phase bone imaging, delayed PET acquisitions, two-day stress and rest protocols all generate studies whose parts arrive separately and must be read together. The routing rules need to hold those parts for one reader rather than scattering them across the panel, which is a worklist-engine capability to demand explicitly and test with a two-day protocol during your pilot.

Camera quality control stays yours. Uniformity and calibration on the gamma cameras, like the rest of the radiation safety program, belongs to the facility. What the reading side owes you is the loop closure: a service whose readers flag QC-degraded studies back to your technologists and chief tech, in writing, catches drift months before an accreditation cycle would. Ask how the vendor handles exactly that report, and require critical findings, an unexpected finding on a bone scan behaves like one on a CT, to be phoned and documented the same way; AstraRad's double-read program samples nuclear studies the same as every other modality, 1 in 20 blind re-read, with major discrepancies under 0.3 percent of signed reports, and its tier clocks are published on the SLA page.

What should you check in a nuclear medicine teleradiology vendor?

The generic vendor checks apply, verifying claims against documents most of all. Layer these modality-specific ones on top.

1. Named readers and their nuclear credentials. Board certification in nuclear radiology or nuclear medicine is publicly verifiable; fellowship files cover the rest. Ask specifically how many readers carry the nuclear volume and what their trailing-year nuclear case count is, because a panel where nuclear is one reader's side interest is a panel with a single point of failure.
2. Final reads, at the same standard as your CT stream. Some services quietly treat nuclear as a preliminary-only modality. Final signed reports should be the default; AstraRad signs finals at every tier on every modality.
3. Scope clarity on cardiac SPECT. In or out, in writing, so nothing falls between your cardiology group and the vendor.
4. The processed-data test during the pilot. Send a renal, a gastric emptying, and a two-day protocol through the pipeline before go-live and read the resulting reports critically. This one exercise surfaces most of what will go wrong in year one.
5. Pricing by study type on a written card. Gamma camera studies and PET-CT are different work and price differently in the market; both belong as printed lines. Per-report pricing with no minimums fits nuclear volume particularly well, because most facilities' nuclear counts are modest and variable, and a minimum sized to your CT stream quietly taxes your nuclear one. AstraRad prices each study type as one figure per signed report, with the priority multiplier printed on the card.

The panel behind the service matters more here than in any common modality, simply because the reader pool is small. Subspecialty composition and counts are on our radiologists.

The line to hold in procurement

Keep the two halves of nuclear medicine separate through every vendor conversation, and the deals become easy to compare. The material half, licenses, authorized users, radiation safety, camera QC, is yours and never moves; the federal framework in 10 CFR Part 35 and its 40-state Agreement State mirror binds the licensee, and no reading contract touches it. The interpretation half moves cleanly to any reader with state licensure, real nuclear training, and the processed data in hand.

AstraRad reads gamma camera and PET-CT studies through fellowship-trained subspecialists, signs final reports at every tier, and prices per signed report with no minimums, no subscriptions, and no platform fees, which suits the modest and variable nuclear volumes most facilities run. Request a rate card with your study types, nuclear lines included, and it reaches you within one business day.