BCV Online Academy: Expert radiology learning. Anytime. Case-based.
NucMed Academy: Nuclear medicine education — interactive and case-driven.
BCV Summer of Cases: Treat yourself to a Summer Case. Check out the blog!
open.cases.med: Share cases. Teach better. Get feedback. Open your account now!
Osteitis Condensans: Understand OCI — and avoid common diagnostic traps. Read the blog.
A chest CT scan can consist of several thousand images. An MRI exam includes different sequence types, weightings, and planes. Sharing just two or three screenshots from these may show selected findings—but not the entire exam.
That is exactly why DICOM exists. The standard preserves not only individual images but also the medical and technical context in which these images were created. And that is precisely why complete, scrollable DICOM datasets play such an important role in clinical collaboration, research, and education.
DICOM stands for Digital Imaging and Communications in Medicine. It is the international standard for the storage, transmission, processing, and display of medical image information. It connects modalities such as CT, MRI, X-ray, or ultrasound with PACS—that is, the image archive—reading workstations, and other medical information systems—even when these come from different manufacturers.
The commonly used description of DICOM as a “wrapper” or “container” is therefore helpful but incomplete. A DICOM file can indeed contain both image data and a wealth of descriptive information. Beyond that, DICOM also defines data models, communication services, and rules governing how systems exchange medical information.
DICOM is not just the image. DICOM is the image, its context, and the instructions on how it must be correctly displayed.
A typical DICOM file contains a structured dataset consisting of individual information elements. These may include, for example:
Added to this are the actual pixel values. In a CT examination, for example, these values determine which measured value corresponds to which pixel, how the slices are spatially oriented relative to one another, and how the CT values to be displayed are calculated from the stored values.
DICOM can also store far more than just traditional individual images. The standard includes, among other things, multiframe data sets, segmentations, structured reports, presentation states, dose information, and objects from radiation therapy.
Typically, a DICOM file does not contain a separate TIFF file, nor is it simply a standard JPEG that could be opened independently of the rest of the data set.
The pixel values are part of the DICOM dataset. They can be stored in a native, uncompressed format or in a so-called “encapsulated format.” A “transfer syntax” describes how the entire dataset—and in particular the pixel information—was encoded.
DICOM supports various compression methods, including:
JPEG-LS and JPEG 2000 can support both lossless and—depending on the transfer syntax—lossy variants. DICOM itself does not specify in which clinical situations lossy compression is acceptable; this must be determined by the respective medical and regulatory context.
There is also no single answer regarding bit depth that applies to all DICOM images.
Among other things, DICOM distinguishes between:
An image can therefore, for example, reserve 16 bits per pixel value but actually use only 12 bits. The remaining bits then contain no additional image information. The DICOM standard explicitly describes such scenarios.
Monochrome medical image data is often stored with more than eight bits, since significantly more gray levels are required than in a typical screen capture. Color images, video recordings, or certain secondary images, on the other hand, may use eight bits per color channel.
It is also important to note that the stored bit depth is not identical to the number of gray levels visible simultaneously on a monitor. Through windowing, a selected range of values is mapped to the displayable brightness levels.
A 16-bit DICOM image does not automatically display 65,536 gray levels simultaneously. It provides a large range of information from which the viewer selects the diagnostically relevant section to display.
Before DICOM became widely established, manufacturers sometimes used their own data formats and communication channels. As a result, the exchange of data between imaging devices, archives, and reading stations was correspondingly complicated.
DICOM creates a common language. As a result, a CT scanner from one manufacturer can generate images that can be archived in a PACS from another manufacturer and viewed on yet another workstation. Among other things, the standard ensures that image sequence, spatial orientation, scale, and essential display information are preserved.
Without this context, a medical image would quickly become nothing more than a collection of pixels.
DICOM data sets are shared for various reasons:
Pre-admission records are shared among hospitals, medical practices, and specialized centers. For a second opinion, it is often not only the written findings that are important, but the complete examination, including all imaging studies.
In tumor boards, rheumatology case conferences, or surgical planning sessions, participants must be able to review the images themselves. Selected screenshots are often insufficient for this purpose.
Research projects often require original image data, imaging parameters, and spatial relationships. Segmentation, quantitative analysis, and the development of AI methods also frequently rely on structured medical image data.
In teaching, a particular advantage is that students can examine the case themselves. They must locate the findings, select the appropriate slice, adjust windowing settings, and evaluate competing abnormalities.
A selected slice can demonstrate a finding. A scrollable dataset shows how to find it.
A textbook image has already been curated. In most cases, the specific section where the pathology is most clearly visible has been selected. Arrows, image cropping, and captions further guide the viewer to the correct location.
In everyday clinical practice, however, the work begins earlier: Which series is relevant? In which plane is the change most clearly visible? Is the finding present on only one slice? Does it extend further? Are there other abnormalities?
Scrollable DICOM cases depict this diagnostic process much more realistically. Learners can navigate through series, zoom, measure, and adjust windowing just as they would on a PACS workstation. Browser-based systems such as PACSbin or Collective Minds Radiology provide complete DICOM cases in a PACS-like environment. The BerlinCaseViewer additionally combines a DICOM viewer with structured cases, annotations, questions, and didactic learning formats.
Viewers such as OHIF, Weasis, Radiant, OsiriX, or Horos can also be used in teaching scenarios. However, they are primarily image viewers or technical frameworks and do not automatically constitute complete teaching platforms with case descriptions, questions, feedback, and learning progress tracking.
DICOM files may contain directly identifying information such as names, dates of birth, patient IDs, or examination numbers. In addition, there are less obvious identifiers: dates, UIDs, free-text fields, information about institutions, staff, or devices, as well as manufacturer-specific private tags.
Personal information may also be embedded directly within the pixels. Examples include names burned into older ultrasound images, scanned documents, or visible labels. In three-dimensional examinations of the head and face, the anatomy itself may even allow for identification under certain circumstances. The DICOM standard therefore provides additional options for burned-in information and recognizable visual features in addition to the basic confidentiality profile.
The standard itself sets forth an important limitation: The use of a DICOM confidentiality profile does not automatically guarantee that an entire information object is completely anonymous. The purpose of the disclosure, the recipient group, the legal basis, and the risk of re-identification must also be taken into account.
De-identification is not a single deletion process. It is a process consisting of data cleansing, risk assessment, and monitoring.
These terms should not be used interchangeably.
In pseudonymization, direct identifiers are replaced, for example, with a code. If it is still possible to link the data back to the original person using additional information, the data remains personal.
Effective anonymization, by contrast, requires that the data subject can no longer be identified by reasonable means. Irreversibly anonymized data is no longer considered personal data under European data protection law; health data that can still be linked to a specific person, however, enjoys special protection.
In practice, this means that simply removing the patient’s name is not sufficient.
The appropriate approach depends on the purpose.
If the images are needed for treatment, they are typically transferred via secure institutional systems—for example, through direct image transfer, a patient portal, a regional network, or a controlled PACS-to-PACS workflow. In such cases, the patient’s identity must often be preserved so that the examination can be correctly assigned.
In these cases, the workflow should include at least the following steps:
Time-limited links help reduce uncontrolled, permanent availability. However, they do not replace proper de-identification or the verification that the case may be shared for the intended purpose.
With open.cases.med, we aim to simplify this process, which has often been cumbersome in the past.
The basic concept is intentionally low-threshold: DICOM images are uploaded, automatically de-identified within the workflow, and then made available directly in the browser as a complete, scrollable case. Instead of sharing large ZIP files or data storage media, the case can be shared via a link.
Depending on the purpose, such a link can be time-limited—for example, for a case conference, a course, or a short-term second opinion. Particularly valuable teaching cases can remain available for a longer period or permanently, provided that rights, data protection, and intended use permit it.
The platform is designed for secure case collaboration, structured expert feedback, and transparent workflows.
open.cases.med does not replace BerlinCaseViewer. The two platforms have different focuses:
Upload. De-identification. Scroll. Share.
The path from PACS to professional exchange shouldn’t be any more complicated than necessary.
For many years, the patient CD was the standard way to share radiological examinations. It was slow, prone to errors, and required that a suitable viewer be installed or included on the disc.
Browser-based DICOM viewers are fundamentally changing this process. Images no longer need to be exported as screenshots, and recipients do not need any special local software. They simply open a link and can view the examination directly within a controlled access environment.
This not only creates a more convenient technical solution; it also changes the way medical cases are discussed and taught—moving away from static images and toward independent exploration.
DICOM is far more than just a file extension. The standard combines image data, technical information, spatial relationships, and medical context. It is only through this integration that individual pixels become an interpretable examination.
However, the dual nature of DICOM becomes particularly evident when it comes to sharing: the extensive context makes the format medically valuable—and at the same time, challenging from a data protection perspective.
The solution, therefore, cannot be to replace DICOM with a few JPEG screenshots. What is needed are tools that make complete image datasets easily accessible while also addressing de-identification, access control, and a clearly defined purpose of use.
Platforms like open.cases.med can help bridge this gap: upload complete cases, process them securely, review them in a browser, and share them selectively with others.
After all, good imaging should not be hindered by file formats, storage media, or technical barriers.
You are currently viewing a placeholder content from Vimeo. To access the actual content, click the button below. Please note that doing so will share data with third-party providers.
More InformationYou are currently viewing a placeholder content from YouTube. To access the actual content, click the button below. Please note that doing so will share data with third-party providers.
More InformationYou need to load content from reCAPTCHA to submit the form. Please note that doing so will share data with third-party providers.
More InformationYou are currently viewing a placeholder content from Facebook. To access the actual content, click the button below. Please note that doing so will share data with third-party providers.
More InformationYou are currently viewing a placeholder content from MailerLite. To access the actual content, click the button below. Please note that doing so will share data with third-party providers.
More InformationYou are currently viewing a placeholder content from Instagram. To access the actual content, click the button below. Please note that doing so will share data with third-party providers.
More Information
