Radiology Systems

What Is PACS in Healthcare?

PACS stands for Picture Archiving and Communication System. It stores, retrieves and displays medical images. This is what it is responsible for, what it is not, and which everyday problems get wrongly blamed on it.

ADThe AbbaDox Team

Updated 18 August 20265 min read

Contents

The short answer

What is PACS in healthcare? PACS stands for Picture Archiving and Communication System. It is the software that stores, retrieves and displays medical images: the CT, MRI, X-ray, ultrasound and mammography studies a radiology department produces. A PACS holds the images themselves. The scheduling, the patient record and the report live in a separate system, usually a RIS.

A corridor of backlit translucent panels receding in perspective, standing for an ordered archive

Key takeaways

PACS is the picture system. If it is not an image, it is probably not the PACS's job.
PACS speaks DICOM, the standard that carries both the image and the data describing it.
Most imaging operations run a PACS and a RIS together. Neither is a substitute for the other.
A surprising share of "PACS problems" are not PACS problems. See below.

What is PACS in healthcare, letter by letter

PACS is an abbreviation for Picture Archiving and Communication System. It is pronounced "packs", and you will see it written several ways that all mean the same thing: PACS, a PACS system, a PAC system, or in the plural, "what are PACS in healthcare". The plural is a misreading rather than a different thing. The S is already part of the abbreviation, so there is no such object as a single "PAC".

Each word in the name is doing real work:

Picture. Medical images from the modalities: CT, MRI, X-ray, ultrasound, mammography, nuclear medicine.
Archiving. Long-term storage at full diagnostic quality, for as long as retention rules require, which is often decades.
Communication. Moving those images to whoever needs them: the radiologist reading them, the referring physician, another facility.
System. Storage, a database, a viewer and the interfaces that connect it to everything else.

What a PACS is made of

Every part below is defined by the DICOM standard, which is what lets a scanner from one manufacturer store to an archive from another. PACS is not a proprietary format; it is an implementation of a published standard.

Whatever a vendor calls its product, a PACS is four things working together. Naming them separately helps, because when something breaks it is nearly always one of the four rather than "the PACS".

The four parts, and what goes wrong in each

Comparison

Part	What it is	What goes wrong here
Modalities	the CT, MRI, X-ray, ultrasound and mammography units producing the studies	the study is labeled wrongly at acquisition
Network	the path between modalities, archive and workstations	"the PACS is slow"
Archive	storage, plus a database of what is stored	retrieval of older studies
Workstations and viewers	diagnostic displays, plus lighter web viewers for everyone else	the wrong person cannot see it

The modalities are not strictly part of the PACS, but nothing downstream works if what they send is wrong, so they belong in any honest diagram of it.

What a PACS is not responsible for

This is the part most explainers skip, and it is the part that causes the most confusion when someone is choosing systems.

A PACS does not schedule the patient. It does not register them, capture their insurance, track them through the department, produce the report, or bill for the study. Those belong to the RIS, and in a hospital some of them belong to the EHR.

Question	Answered by
When is the patient coming, and for what?	RIS
Who is the patient, and what is their history?	RIS, or the EHR in a hospital
Where is the image?	PACS
What does the image show?	PACS, read by a radiologist
What did the radiologist conclude?	RIS
Who gets the report, and when?	RIS
What do we bill?	RIS, or the billing system

What people wrongly blame on the PACS

We see this constantly from the other side of the integration, and it is worth naming because it changes what you should actually go and fix.

"Priors are missing." Usually the PACS holds them. The study did not match because the patient identifier or accession number differed between systems, so the viewer never found them. That is an interface problem, not a storage problem.
"Images are slow to load." Sometimes the PACS. Often the network between the reading station and the archive, or a storage tier that has moved older studies to slower media.
"The wrong study opened." Almost always a worklist problem: the order the modality received did not carry the right detail, so the acquisition was labeled incorrectly at the source.
"The referring physician cannot see the images." Usually a distribution and permission question rather than the archive.

The pattern: when images exist but the right person cannot see the right one at the right time, look at the handoffs before you look at the PACS.

How a PACS fits with everything else

An order arrives and the RIS schedules it. The modality picks up the work from a worklist. The study is acquired and sent to the PACS. The radiologist opens it, with the RIS supplying the clinical context and the priors. The report goes back out through the RIS.

The images travel on DICOM. The messages about patients, orders and results travel on HL7, increasingly FHIR. When someone says two systems are integrated, those are the standards they mean, and the depth of that integration varies enormously between vendors.