Dermatology image management workflows are usually imagined as a single upload button inside the EHR. In practice, the workflow spans at least four connected decisions: what standard governs the file and its metadata, how a regional or overview photograph is linked to dermoscopic images of the same lesion, where the object is archived, and whether the EHR mobile app is allowed to participate. The DICOM Supplement 221 specification supplies the most specific technical model for dermoscopy, but published implementations describe a different, parallel path built around smart-device photodocumentation. A dermatology group planning an image workflow therefore needs to decide early whether it is building a standards-based DICOM archive for longitudinal lesion tracking or a lighter EHR-integrated documentation program, because downstream retrieval, metadata, and interoperability differ materially between them.
What Supplement 221 Names and Why the Linkage Model Matters
DICOM Supplement 221 introduces a Dermoscopic Photography Image Storage SOP Class, with UID 1.2.840.10008.5.1.4.1.1.77.1.7, and a corresponding information object definition for encoding and storing dermoscopic images [1]. The supplement also defines a recursive relationship type for one interpretive dermatology use case: a regional image may contain multiple lesions, a lesion may appear in multiple regional images, and therefore the regional-to-dermoscopic relationship is many-to-many [1]. A technician may photograph an entire forearm, then acquire close dermoscopic views of several moles on that forearm and bind those close-ups to the regional view.
The value of this model is not just organizational. The supplement suggests using the skin lesion identifier as the series descriptor for dermoscopic images of a given lesion, with the regional image metadata carrying referenced SOP Instance UIDs [1]. For longitudinal comparison, it proposes a lesion tracking workflow in which a tracking identifier stores the skin lesion label and a tracking unique identifier stores a vendor-generated lesion UID; comparison requires images from the same modality [1]. These mechanisms address a real operational failure: dermatology photos stored as unnamed JPEGs in a folder cannot be reliably compared with a future image of the same mole because the lesion identity and modality context were never recorded.
The Photodocumentation Alternative: 31,897 Images Across 54 Specialties
A separate implementation report from a large healthcare organization describes a smart-device photodocumentation program that uploaded 31,897 photo-based imaging studies from November 13, 2018 through December 31, 2023 across 54 medical specialties [2]. The authors frame the project explicitly as a response to integration challenges between medical photography, the EHR, and standard clinical workflow [2]. This is a different problem framing from the DICOM dermoscopy supplement: the photodocumentation program sought broad clinical documentation utility, not lesion-specific longitudinal tracking.
The efficiency comparison inside that report is instructive. A smart-device capture-and-send workflow required a counted number of clicks, taps, or swipes for a non-first-time user already logged into the patient chart, while a DSLR capture and upload workflow required at least 10 clicks and keystrokes [2]. The authors add that the DSLR count is likely undercounted because users may need to search a computer folder structure or transfer images from camera storage [2][8]. This is not a vendor benchmark; it is one organization's measurement of its own two workflows. Still, it shows that image management design is partly a time-motion decision, not merely a standards decision.
A Minimal Workflow Specification for Evaluation
As a planning artifact, a dermatology group could draft a four-line specification before requesting demos. First, images acquired for longitudinal lesion comparison must be encoded and stored in the Dermoscopic Photography Image Storage SOP Class, not as an attachment-only or consumer JPEG [1]. Second, regional images must hold referenced dermoscopic SOP Instance UIDs and a skin lesion identifier so lesion identity survives archive migration [1][6]. Third, mobile EHR capture is permitted only if photographs can be sent to the imaging archive and receive medical imaging metadata; otherwise the group should explicitly accept non-comparable photos [2][8]. Fourth, the group should measure click or keystroke counts for each capture path during a timed test and record any hidden manual steps such as folder selection or camera media transfer [2][8].
How Images Are Actually Stored Today
International survey data offer a wider snapshot of current dermatology image handling. The authors distributed a 22-question survey through the International Dermoscopy Society website and ISIC email listserv between October 6, 2023 and February 6, 2024, collecting 228 responses; most respondents practiced in Europe (46.56%) in independent (42.02%) or academic (25.46%) settings [3]. The most common storage methods were personal devices (45.09%), secure software (28.60%), and office devices (24.63%) [3]. Those are methods distributed across the group, not a single practice’s default.
This distribution has a direct consequence for central archive planning. An analysis of dermatologic imaging metadata notes that dermatologic storage is often distributed across acquisition modalities or department-level systems, which inhibits a patient-centric view of all imaging [6]. The same analysis argues that a centralized repository lets treating physicians see all patient images and lets departments such as histopathology view them [6]. An enterprise imaging strategy, the authors note, may reduce cost, scalability, management, reliability, availability, and security concerns, though that is a design position rather than a measured outcome [6].
Technical Decision Points: Archive, Metadata, and Mobile Capture
Two technical decisions recur across the sources. The first is DICOM versus consumer image file formats. The metadata review states that DICOM provides a comprehensive, international medical image management standard with standardized, structured, and coded metadata models that are patient-, modality-, and specialty-specific [6]. The same paper says DICOM appears to provide an advantage over consumer formats for metadata because its model includes the patient, study, and technical metadata needed for clinical use [6]. That advantage is conditional on the receiving systems actually validating and using those fields.
The second decision is mobile EHR upload. The photodocumentation program deliberately did not allow image upload through mobile EHR applications because those photographs could not be sent to the imaging archive, could not be compared over time, and could not receive medical imaging metadata [2][8]. The authors judged those deficiencies to outweigh the benefit of inserting a photograph directly into a note [2]. This is a governance choice, not a universal prohibition; other organizations may accept the archive gap for convenience. But the tradeoff should be named in any workflow requirement sheet.
An Evaluation Template for a Dermatology Group
A useful evaluation can be staged as a documentation to archive walkthrough rather than a feature checklist. The suggested first scenario: a nurse photographs the left forearm with a regional image, then captures dermoscopic views of two lesions and links both close-ups to the regional image using the skin lesion identifier as series descriptor [1]. The evaluation should record whether the system can store the regional-to-dermoscopic relationship as referenced SOP Instance UIDs, or whether the link exists only in the photographer's memory [1].
The suggested second scenario: the same patient returns eight months later. The dermatologist retrieves the original dermoscopic image of a suspicious mole and opens a tracking view. The evaluator checks whether the system can use a tracking identifier and tracking unique identifier to recall the prior image from the same modality, as the supplement describes [1]. The third scenario concerns non-standard capture: a resident uses a DSLR and manually transfers files. The evaluation should count the actual clicks and keystrokes and note any folder navigation or camera-media removal steps, following the measurement approach used in the photodocumentation assessment [2][8].
Sources & references
- Digital Imaging and Communications in Medicine (DICOM) · www.dicomstandard.org
- Implementing a Photodocumentation Program - PMC - NIH · pmc.ncbi.nlm.nih.gov
- International Survey on Dermoscopic Image Management: ISIC Data on Capture, Storage, and AI Integration in Dermatology - PMC · pmc.ncbi.nlm.nih.gov
- Joint VA / DoD DICOM Conformance Requirements for Digital Acquisition Modalities · www.va.gov
- Electronic Health Record (EHR) IHS Imaging Viewer and the IHS all Images Viewer · www.ihs.gov
- Transforming Dermatologic Imaging for the Digital Era: Metadata and Standards - PMC · pmc.ncbi.nlm.nih.gov
- Interactive Journal of Medical Research - Recommendations for Better Adoption of Medical Photography as a Clinical Tool · www.i-jmr.org
- Implementing a Photodocumentation Program | Journal of Imaging Informatics in Medicine | Springer Nature Link · link.springer.com