Artificial intelligence paired with three-dimensional total-body photography is moving skin cancer surveillance toward automated, whole-body lesion tracking, according to a review published in Frontiers in Medicine.

The paper examines the convergence of two technologies that dermatology clinics have so far adopted separately: 3D total-body photographic systems, which capture a rotating matrix of images of the entire skin surface, and machine-learning algorithms trained to detect and match pigmented lesions across serial scans. Together, they allow clinicians to follow every visible lesion over time rather than rely on isolated dermoscopic snapshots or patient-reported changes.

That longitudinal, whole-body framing matters clinically. Most melanomas arise de novo rather than from pre-existing nevi, and detecting a new or subtly changing lesion among hundreds is precisely the task where human surveillance strains. Total-body photography has long addressed this problem in analog form for high-risk patients, but manual comparison of hundreds of lesion images across visits is time-consuming and error-prone. The Frontiers review frames AI as the component that makes 3D whole-body imaging scalable in routine practice.

The technical pipeline works roughly as follows. The 3D imaging system constructs an avatar of the patient's skin surface; computer-vision algorithms then register lesions between visits, flag new lesions, and score detected changes — in size, shape, or color — for follow-up priority. Because every scan maps the same body surface in three dimensions, lesion matching does not depend on the patient assuming an identical pose or the photographer replicating an identical angle, a known limitation of conventional two-dimensional total-body photography.

The review situates the technology within the broader skin-cancer screening problem. Incidence of both melanoma and non-melanoma skin cancer continues to rise, and dermatologist capacity has not kept pace. AI-assisted triage could shorten the interval between a suspicious change and its clinical assessment, and could reduce unnecessary excisions of benign lesions that look static when compared against prior imaging.

Commercial 3D total-body photography systems have already entered dermatology practices, and convolutional neural networks have demonstrated dermatologist-level performance in controlled image-classification studies of skin lesions. The review's contribution is to assess how these strands combine at the level of the surveillance workflow: capture, registration, change detection, and clinical decision support.

The authors also weigh the constraints. Algorithm performance depends heavily on the quality and diversity of training data, and datasets skewed toward lighter skin tones raise equity concerns for patients with skin of color — a documented limitation across dermatologic AI. Regulatory pathways for AI-based diagnostic support, integration with electronic health records, reimbursement for imaging-based surveillance, and the medico-legal status of AI-flagged findings all remain unsettled. The review argues these questions must be resolved before whole-body AI surveillance can move from specialized melanoma centers to general dermatology.

For dermatologists managing patients with numerous nevi, personal or family history of melanoma, or genetic predisposition syndromes, the technology promises a standardized, reproducible baseline against which every future lesion can be compared. For health systems, it hints at a future where screening capacity scales with computing rather than with clinician hours.

The review concludes that AI-enabled 3D total-body photography is among the most mature applications of artificial intelligence in clinical dermatology, and that prospective studies comparing AI-assisted surveillance with standard care on melanoma detection rates and excision ratios will determine whether it earns a permanent place in the skin-cancer pathway.