(0028,0302)
Recognizable Visual Features
Keyword: RecognizableVisualFeatures
Indicates whether the image contains patient-recognizable visual features (e.g., facial structures on head CT/MR).
Tag Attributes
- Tag ID
- (0028,0302)
- Compact
- 00280302
- Keyword
- RecognizableVisualFeatures
- VR
- CS — Code String
- VM
- 1
- Required
- Type 3 — optional
- PS3.15
- Not listed in PS3.15 Table E.1-1
- Retired
- No
Defined in
What Recognizable Visual Features contains
Indicates whether the image contains features, such as a face reconstructed from a head CT or MR, that could identify the patient. It matters when sharing data for research.
Group 0028 describes the pixel matrix and how stored pixel values are turned into displayed values.
Value format: CS (Code String), VM 1
(0028,0302) RecognizableVisualFeatures has Value Representation CS (Code String) and Value Multiplicity 1, which means exactly one value.
CS — Code String
A short code. Leading and trailing spaces are not significant. Codes are usually Defined Terms or Enumerated Values listed by the standard.
- Length
- 16 bytes maximum
- Characters
- Uppercase letters A–Z, digits 0–9, the space and the underscore.
Example value
NO
Format mistakes to avoid
- Lowercase letters, hyphens and other punctuation are not allowed, so ct or Head-Neck are invalid code strings.
- Values longer than 16 characters are invalid.
Common values
| Value | Meaning |
|---|---|
| YES | Visual features can identify the patient |
| NO | No patient-identifiable visual features |
Where RecognizableVisualFeatures appears
| Defined in | Type | Meaning |
|---|---|---|
| General Image ModulePS3.3 C.7.6.1 | 3 | Optional. |
A module lists the attributes an object must or may carry. Many IODs (CT Image, MR Image, Secondary Capture and others) include the same module, and an IOD-specific module can make an attribute stricter than the general one.
De-identification (PS3.15)
Recognizable Visual Features is not listed in PS3.15 Table E.1-1, so the Basic Application Level Confidentiality Profile gives no specific action for it.
Recognisable faces in head CT and MR are a pixel-data problem, so removing attributes does not help. PS3.15 has a Clean Recognizable Visual Features Option for this case.
To see how de-identification changes your own files, try the free DICOM Anonymizer, which runs in the browser and uploads nothing.
In archives and migrations
Image pixel attributes must stay consistent with the Pixel Data. If images are recompressed during a move (for example to JPEG 2000), the transfer syntax changes and some of these values can change with it, so a lossless copy is the only way to keep them byte-for-byte identical.
Common errors and how to fix them
Troubleshooting: Before sharing head CT or MR for research, check this attribute and do not trust its absence: it is optional. Defacing or skull-stripping is done on the pixels, outside the scope of attribute de-identification.
Check a file in seconds with the free DICOM Validator: it reads the header in your browser, lists the tags with their values, and flags missing or malformed key attributes.
Recognizable Visual Features: frequently asked questions
What is DICOM tag (0028,0302) Recognizable Visual Features?
Indicates whether the image contains patient-recognizable visual features (e.g., facial structures on head CT/MR).
What VR and VM does Recognizable Visual Features (0028,0302) have?
(0028,0302) RecognizableVisualFeatures has Value Representation CS (Code String) and Value Multiplicity 1, which means exactly one value. A short code. Leading and trailing spaces are not significant. Codes are usually Defined Terms or Enumerated Values listed by the standard. Maximum length: 16 bytes maximum.
Is Recognizable Visual Features (0028,0302) required?
In the General Image Module (PS3.3 C.7.6.1) it is Type 3: optional.
How is Recognizable Visual Features handled when de-identifying DICOM files?
Recognizable Visual Features is not listed in PS3.15 Table E.1-1, so the Basic Application Level Confidentiality Profile gives no specific action for it. Recognisable faces in head CT and MR are a pixel-data problem, so removing attributes does not help. PS3.15 has a Clean Recognizable Visual Features Option for this case.
How do I troubleshoot Recognizable Visual Features (0028,0302)?
Before sharing head CT or MR for research, check this attribute and do not trust its absence: it is optional. Defacing or skull-stripping is done on the pixels, outside the scope of attribute de-identification.
Check RecognizableVisualFeatures in your files
The free DICOM Validator lists every header tag and flags problems. Files stay in your browser.
Open DICOM Validator →Remove patient data before sharing
The free DICOM Anonymizer applies a de-identification profile in your browser, with no upload.
Open DICOM Anonymizer →Sources
Checked against the DICOM standard, release 2026d. Module Types, query keys and de-identification actions come from these tables: