Total-Body Photography (mole mapping)

What it does

Total-Body Photography reads a mole-mapping session: a 3D total-body capture or a 2D whole-body photo series of a patient under melanoma surveillance. The run detects the pigmented and non-pigmented lesions visible on the skin, places each one on a body map with its own lesion identity, and counts the naevi.

It then ranks the lesions within the patient: a lesion that looks unlike the patient's other moles (the "ugly duckling" sign) comes first. Each lesion also gets a malignancy score. With a prior session it matches lesions across the two captures and lists the new, enlarged and changed ones. The read ends with the short list of lesions that should be examined with a dermatoscope.

Intended use

Decision support for surveillance of high-risk patients: those with many or atypical naevi, a previous melanoma or a strong family history. A full session holds hundreds of lesions, and the run lists the few that should be examined with dermoscopy first. A high score sends the lesion to dermoscopy (Dermoscopy · Skin Lesion) for a read at full magnification.

Who it is for

Dermatologists, pigmented-lesion and high-risk melanoma clinics, and trained mole-mapping nurses. Developers can feed session exports from total-body imaging systems into the run and store lesion identities for longitudinal follow-up.

Accepted input

• 3D total-body captures: vendor exports (for example VECTRA WB360) as mesh with texture, or as 2D tiles.
• 2D whole-body photo series: standardised photographs of each body region from one session.
• Prior session: the previous capture of the same patient in prior_files for the change section.

Requirements

Requirement	Why
A complete session	The within-patient ranking compares each lesion with all the others
Consistent poses and regions between sessions	New and changed lesions are found by matching body-map positions
Patient age and history	Part of the per-lesion score context

Optional context

Age, sex, personal or family history of melanoma, previous excisions and patient-reported changing lesions can be given in clinical_context.

The result

One JSON result with a lesion list: each lesion has a stable lesion_id, body-map coordinates, a box on the image and its scores.

Section	Content
Lesion map	All detected lesions with body-map coordinates and boxes, and the naevus count
Suspicious lesions	Per-lesion malignancy score and the within-patient ranking, with the dermoscopy targets
New and changed lesions	Lesions that are new, enlarged or changed since the prior session
Draft report	Opt-in (options.report): an English narrative that interprets the result for the reader (findings, impression, limitations); every number is checked against the findings

Standards

• Lesion identity: FHIR BodyStructure per lesion, so the same lesion is found again at the next session.
• Body site: DICOM CID 4029 Dermatology Anatomic Site codes with 2D or 3D body-map coordinates.
• Exports: FHIR DiagnosticReport and Observation for counts and scores; Key Object Selection for the images of the dermoscopy targets; Encapsulated PDF.

Working with Dermoscopy

The suspicious-lesion list is the hand-off to a separate dermoscopy request for each flagged lesion, where structures, a differential and a management class are read at full magnification.