rel="preload" href="https://fonts.googleapis.com/css2?family=Inter:wght@400;500;600;700;800&display=swap" as="style" onload="this.onload=null;this.rel='stylesheet'" />
As a radiologic technologist, your primary focus is on positioning, exposure factors, and patient care. But every image you produce passes through a sophisticated digital ecosystem long before it reaches the radiologist's reading station. Understanding how PACS and DICOM work isn't just an IT concern — it is increasingly tested on the ARRT exam and essential for troubleshooting your daily workflow.
Whether you work in a large hospital system with a fully integrated enterprise PACS or a small clinic with a mini-PACS attached to a single DR unit, the principles are the same. Your images must be acquired, formatted, transmitted, stored, retrieved, and displayed correctly. When a study fails to appear on the radiologist's worklist, the technologist who understands the system can often solve the problem before it requires an IT ticket.
This guide covers the PACS and DICOM fundamentals that every rad tech needs, from understanding image tags and compression to navigating RIS-PACS integration and troubleshooting common connectivity problems. Let this be your go-to reference for the digital side of radiography.
The ARRT exam includes questions on image data management, DICOM standards, information systems integration, and quality assurance of digital image handling. Expect 3–5 questions related to PACS, DICOM, and RIS on your registry exam. Understanding these concepts also affects the "Image Production" and "Patient Care" content categories.
PACS stands for Picture Archiving and Communication System. It is the hardware and software infrastructure that enables digital medical images to be acquired from modalities, stored on servers, transmitted across networks, and displayed on workstations — all without film.
Every PACS has three essential subsystems:
The gateway that receives images from modalities (CT, MR, DR, CR, US, NM, etc.) via DICOM send. Includes modality worklist (MWL) integration to pre-populate patient data.
Short-term RAID storage for active studies (30–90 days) and long-term archive (NAS, SAN, or cloud server) for permanent storage. Includes hierarchical storage management (HSM).
Diagnostic-quality monitors for radiologists and review stations for clinicians. Includes image manipulation tools (window/level, zoom, pan, measurements, MPR for CT/MR).
These three components are connected by a DICOM network that handles communication between all devices. Most modern PACS also include a web-based viewer (VNA or zero-footprint viewer) that allows clinicians to access images from any hospital computer without installing special software.
Not all PACS are created equal. You will encounter different configurations depending on the size of your facility:
When starting at a new facility, learn the PACS workflow within your first week. Ask your lead technologist or PACS administrator: (1) How do I check that a study was sent successfully? (2) How do I handle a "send failure" message on the modality console? (3) What is the process for query/retrieve if a study needs to be re-pulled from the archive? Knowing these three things will save you time and prevent repeat exams.
DICOM (Digital Imaging and Communications in Medicine) is the international standard (ISO 12052) that defines how medical images are formatted, stored, transmitted, and displayed. Without DICOM, a CT scanner from GE could not send images to a PACS from Philips or a workstation from Siemens. DICOM is the universal language that makes multi-vendor interoperability possible.
Every DICOM-compliant device — whether it is a CT scanner, DR detector, CR reader, ultrasound machine, MRI scanner, or nuclear medicine gamma camera — implements a specific set of DICOM services. The most important ones for rad techs are:
DICOM organizes image data in a hierarchical model with four levels. Understanding this hierarchy is critical for navigating any PACS:
| Level | Description | Example DICOM Tags | Relationship |
|---|---|---|---|
| Patient | The person being imaged | Patient Name (0010,0010), Patient ID (0010,0020), DOB (0010,0030) | Top level; one patient can have many studies |
| Study | A single imaging exam (one accession number) | Study UID (0020,000D), Study Date (0008,0020), Accession Number (0008,0050) | Unique per exam; triggered by a single order |
| Series | A single scan sequence or acquisition within a study | Series UID (0020,000E), Modality (0008,0060), Series Number (0020,0011) | One study can have many series (e.g., scout, axial, coronal) |
| Image | A single image frame or slice | SOP Instance UID (0008,0018), Image Number (0020,0013), Pixel Data (7FE0,0010) | One series can contain many individual images |
Each of these levels is identified by a globally unique UID (Unique Identifier). The Study UID, Series UID, and SOP Instance UID (for each image) ensure that every image in the world can be uniquely referenced — critical for multi-site health information exchanges and teleradiology.
Every DICOM file contains two parts: the pixel data (the actual image) and metadata stored as DICOM tags. A tag is a structured key-value pair identified by a group number and element number in hexadecimal format. For example, the patient's name is stored at tag (0010,0010), and the modality type is at tag (0008,0060).
Here are the most common DICOM tags you should know for the ARRT exam and clinical practice:
| Tag (Group,Element) | Name | Content Example |
|---|---|---|
| (0008,0020) | Study Date | 20260626 |
| (0008,0030) | Study Time | 143022.500 |
| (0008,0050) | Accession Number | A20260626001 |
| (0008,0060) | Modality | CT, MR, DX, CR, US, NM |
| (0008,0090) | Referring Physician | Smith^John |
| (0010,0010) | Patient Name | Doe^Jane |
| (0010,0020) | Patient ID | MRN987654 |
| (0010,0030) | Patient Birth Date | 19900415 |
| (0010,0040) | Patient Sex | F |
| (0020,000D) | Study Instance UID | 1.2.840.113619.2.55.3... (unique global ID) |
| (0020,0010) | Study ID | 12345 |
| (0028,0010) | Rows (image height) | 512 |
| (0028,0011) | Columns (image width) | 512 |
| (0028,0030) | Pixel Spacing | 0.625\0.625 |
| (0028,1050) | Window Center | 40 (for CT soft tissue window) |
| (0028,1051) | Window Width | 400 |
You will not need to memorize DICOM tag numbers for the registry. However, you should understand the four-level information model (Patient → Study → Series → Image), what the Accession Number represents (the unique identifier for a radiology order), how Modality Worklist reduces data entry errors, and the difference between lossy and lossless compression. These are the most commonly tested DICOM/PACS concepts on the ARRT exam.
In most imaging departments, the PACS and RIS (Radiology Information System) work together seamlessly. The RIS handles the administrative and workflow side: patient scheduling, exam ordering, billing, report transcription, and result distribution. The PACS handles the image side. The two systems communicate through HL7 (Health Level 7) for text-based data and DICOM for image data.
Here is how a typical exam flows from order to report:
If your modality console shows "Study not found on worklist", do not circumvent the system by entering patient data manually — that is how wrong-patient errors happen. Instead, verify that: (1) the exam was properly ordered and scheduled in the RIS, (2) the modality is connected to the correct MWL server, and (3) the scheduled date/time matches today. If the order was placed but does not appear, ask your lead tech or PACS admin to check the HL7 interface between the RIS and MWL broker.
Medical images consume enormous amounts of storage space. A single CT chest study with 1 mm slices can contain 300–500 images at 512×512 pixels each — roughly 150–250 MB of uncompressed data. A busy CT department performing 40–50 scans per day generates 6–12 GB of new data daily. Compression is essential for managing storage costs and network bandwidth.
DICOM supports two types of compression:
| Compression Type | How It Works | File Size Reduction | Used For |
|---|---|---|---|
| Lossless | No image data is discarded. Original pixels can be perfectly reconstructed. | 2:1 to 3:1 (50–67% reduction) | Primary diagnostic interpretation, short-term archive, legal records |
| Lossy | Non-diagnostically-relevant pixel data is discarded (usually high-frequency noise). | 10:1 to 20:1 (90–95% reduction) | Long-term archive, web viewers, teaching files, non-diagnostic review |
JPEG Lossless (DICOM Transfer Syntax 1.2.840.10008.1.2.4.70) is the most common lossless compression in PACS. JPEG 2000 offers both lossless and lossy modes with improved compression ratios. The DICOM standard mandates that every modality and PACS must support at least one lossless compression scheme. Lossy compression should never be applied to images that will be used for primary diagnosis — it is reserved for long-term archives or non-clinical purposes.
Even the most robust PACS encounters problems. Here are the most common issues rad techs face and how to resolve them:
This is the most frequent PACS problem. The modality attempted to send images via DICOM C-STORE but the PACS did not acknowledge receipt. Checklist: (1) Is the network cable connected? (2) Is the PACS destination IP address and port correct in the modality configuration? (3) Is the PACS archive online and accepting connections? (4) Does the PACS have sufficient storage space? Most modalities have a "Send Queue" or "Retry" function — use it before re-scanning the patient.
The images were sent successfully but did not arrive at the reading workstation. Checklist: (1) Verify in the PACS archive that the study was received (most PACS have a "Study Status" search). (2) Check whether the study was "split" or "merged" incorrectly. (3) Ensure the study status is not "Hold" or "Unverified." (4) Verify that the reading radiologist's worklist filter is correct — sometimes the filter excludes certain exam types or time ranges.
Another patient's data appears on your study. This is a serious safety event and must be reported immediately. Most PACS and modalities allow demographic correction — but only by authorized personnel. Never modify patient data on the modality without following your facility's correction policy. If images were sent with incorrect demographics, the study must be deleted from PACS and re-sent with the corrected data.
Images load slowly on the workstation, especially for large CT or MR studies. Check: (1) Network bandwidth — large studies can saturate a 100 Mbps connection. Hospital IT departments should prioritize imaging traffic via QoS (Quality of Service). (2) Pre-fetching settings — the PACS should pre-fetch prior exams for comparison before the radiologist opens a study. (3) Workstation hardware — diagnostic monitors require dedicated graphics cards with sufficient VRAM. (4) Archive type — studies on slower spinning-disk archives take longer to retrieve than those on solid-state storage.
Keep a log of network or PACS errors in your department. Note the date, time, modality, and error message. This data is invaluable for PACS administrators and IT teams when tracking recurring problems. If the same modality produces "send failed" errors daily at 9 AM, there may be a scheduled network task (like a backup) causing a brief interruption. Pattern recognition solves PACS problems faster than any single reboot.
These terms are the most likely to appear on your registry exam related to PACS and DICOM:
Try these ARRT-style multiple choice questions based on this article. Click an option to check your answer — correct answers turn green, wrong ones turn red.