What is dermoscopy? A clinician's guide to the technique, its structures, and digital follow-up

Two pigmented spots can look identical to the naked eye and be nothing alike underneath — one a harmless mole, the other an early melanoma. That gap between what the surface shows and what the skin is actually doing is the problem dermoscopy was invented to solve. By flattening the glare of the outer skin and magnifying what lies just below it, a dermatoscope turns a flat brown spot into a landscape of colors, networks, and vessels that carry real diagnostic meaning. This guide explains what dermoscopy is, how the light works, how clinicians actually read what they see, where the technique reaches far beyond moles, and why capturing it digitally — so the same lesion can be compared visit to visit — is what turns a good look into a durable record.

What is dermoscopy?

Dermoscopy — also called dermatoscopy or epiluminescence microscopy (the three are synonyms) — is a non-invasive technique for examining the skin with a handheld instrument called a dermatoscope: a magnifier, typically around ×10, combined with a controlled light source.^[9] Its whole trick is optical. The outer layer of the skin (the stratum corneum) scatters light and acts like frosted glass; the dermatoscope defeats that scattering — historically with a film of fluid that renders the epidermis translucent — so the observer can see structures within the epidermis, the dermo-epidermal junction, and the papillary dermis that are simply not visible to the unaided eye.^[9] The foundational description came in 1987, when Pehamberger and colleagues examined more than 3,000 lesions and defined a set of morphologic criteria "not readily apparent to the naked eye" — opening what they called a new dimension of skin morphology.^[1]

In practice the examiner is reading two things: colors (which roughly encode where the pigment sits — black high in the epidermis, brown at the junction, gray/blue in the dermis) and structures (pigment network, dots and globules, streaks, vessels, and so on). The point is not the magnification itself; it is that these sub-surface patterns behave differently in benign and malignant lesions.

Why isn't the naked eye enough?

Because the numbers say so. Across studies, adding dermoscopy to clinical examination measurably improves melanoma detection. A meta-analysis of studies performed in real clinical settings found that sensitivity for melanoma rose from 71% with naked-eye examination to 90% with dermoscopy — without a meaningful loss of specificity.^[3] An earlier meta-analysis of 27 studies put the improvement in diagnostic accuracy at roughly 49% over inspection with the unaided eye.^[2] Beyond raw accuracy, trained dermoscopic examination is associated with fewer unnecessary excisions and referrals, because benign lesions can be recognized with more confidence rather than removed "just in case."^[9]

There is an honest caveat that every serious source repeats, and we will not bury it: the benefit depends on training. Kittler's meta-analysis found that dermoscopy improved accuracy only for experienced examiners; in untrained hands, the extra structures can add noise rather than signal.^[2] Dermoscopy is a skill, not a filter you switch on.

How do clinicians read a dermoscopic image?

The first widely adopted method was pattern analysis — reading the overall architecture of colors and structures, introduced by Pehamberger in 1987.^[1] As the field grew, more structured, checklist-style methods were developed to make the reading teachable and reproducible:

• The ABCD rule of dermatoscopy — scores Asymmetry, Border sharpness, Color variety, and Differential structures into a total score.^[6]
• The 7-point checklist — weights seven melanoma-specific criteria; in its revised, simplified form, the presence of at least one criterion flags a lesion for closer scrutiny, favoring sensitivity.^[5]
• Menzies method and the CASH and three-point algorithms — alternative scoring schemes tuned for different settings and experience levels.^[6]

These were compared head-to-head in the landmark Consensus Net Meeting on Dermoscopy, where 40 experienced dermoscopists evaluated 108 lesions over the Internet. The structure that mattered most was the two-step approach — first decide whether a lesion is melanocytic at all, then apply a melanoma algorithm — and, among the second-step methods, classic pattern analysis gave the best diagnostic performance, with the checklist algorithms offering comparable sensitivity but lower specificity.^[4] The practical reading of that result: the algorithms are scaffolding for learners and a safety net for non-experts, but they approximate the holistic pattern recognition an experienced eye performs directly.

Polarized or immersion? Two ways to light the skin

Modern dermatoscopes defeat surface glare in one of two ways, and the difference is not cosmetic — each mode reveals structures the other hides:^[7]

• Non-polarized (immersion contact) dermoscopy presses the lens to the skin over a fluid (alcohol, gel, or oil). It excels at superficial features — milia-like cysts and the blue-white veil show up better this way.
• Polarized dermoscopy uses cross-polarized light and needs no contact and no fluid, which makes it faster and more hygienic. It reaches deeper, showing vessels and pink/red tones more clearly, and it uniquely reveals shiny white structures (also called crystalline or chrysalis lines) tied to altered collagen — a clue polarized light alone can show.^[7]

Agreement between the two modes on the overall pattern is high, but the exceptions are diagnostically important — which is why many dermatoscopes now toggle between polarized and non-polarized in a single device, and experienced users switch between them on the same lesion.^[7]

Beyond melanoma: dermoscopy of (almost) everything

Dermoscopy was born from pigmented lesions and melanoma, but it did not stay there. Over the past two decades its use has expanded across dermatology, to the point that the field half-jokingly calls it "dermoscopy of everything."^[9]

• Non-melanoma skin cancer. Basal cell carcinoma has a well-defined dermoscopic vocabulary — arborizing vessels, blue-gray ovoid nests, leaf-like areas, spoke-wheel structures — that helps distinguish subtypes, separate pigmented from non-pigmented variants, and even assess surgical margins.^[10]
• Inflammatory and infectious skin disease ("inflammoscopy" and "entomodermoscopy") — vascular and scale patterns aid diagnosis and treatment monitoring.^[9]
• Nails. Dermoscopy of the nail bed and matrix helps evaluate pigmented nail bands and other nail disorders.^[14]
• Hair and scalp. Dermoscopy of hair-bearing skin has its own name — trichoscopy, a term coined by Rudnicka and Olszewska in 2006 — and it is now a first-line tool in the workup of alopecias.^[11] (It is the sibling technique to everything on this page; see our guide to trichoscopy.)

The common thread is that the same optical principle — see below the surface, read colors and structures — generalizes far past the mole it started with.

From a single look to a series: digital dermoscopy

Classic dermoscopy happens through the eyepiece and lives in the examiner's memory and notes. Digital dermoscopy attaches a camera, so the image can be stored, measured, and — most importantly — compared later. That last capability changes what the technique can do.

For patients at high melanoma risk, the standard of care combines total-body photography (to catch new or changed lesions across the whole skin) with sequential digital dermoscopy of individual suspicious lesions — the "two-step method of digital follow-up." Watching a flat, featureless atypical lesion over time lets clinicians catch subtle change that no single snapshot would reveal, detecting melanomas earlier and thinner while excising fewer benign lesions.^[8] The whole approach only works if the same lesion can be found and re-photographed the same way months later — which is exactly where standardized capture and clinical software earn their place, and where the technique meets the same standardization discipline as any clinical photography.^[15]

Where AI fits — and where it doesn't (yet)

Because dermoscopic images are structured and visual, they are fertile ground for machine learning. Convolutional neural networks have, in controlled reader studies, matched or exceeded dermatologists at classifying melanoma versus nevus on dermoscopic images.^[12] Used as an assistant rather than an oracle, AI support has been shown to lift clinicians' own sensitivity — in one web-based reader study, from roughly 59% to 75% — while leaving specificity broadly intact.^[13]

Two honesty guardrails belong here, and we hold to them:

1. Retrospective image benchmarks are not the clinic. A model that shines on a curated holdout set still has to prove itself prospectively, across skin types, devices, and the messy lesions that never make it into datasets. "Superior on a test set" is a real result and an incomplete one.
2. Automated output supports the clinician; it does not replace the read — or the diagnosis. A dermoscopic classification is a decision-support signal, and in most jurisdictions a tool that outputs a malignancy verdict is a regulated medical device. On this site, any AI-based lesion analysis appears as a clearly labeled upcoming capability, never as a diagnostic claim.

How the Imaginos platform fits in

You do not need us to do dermoscopy — a dermatoscope and a trained eye do that. Where software matters is the part dermoscopy exposes but does not solve on its own: keeping the image findable, standardized, and comparable long after the appointment. Here is how we think about it.

• Regions organize the record. A lesion or scalp area is documented against a stable anatomical region, so "this spot over the last year" is something you open, not a folder you dig through.
• A digital marker replaces the tattoo. Sequential monitoring has always depended on returning to the exact same spot; the classic answer was to tattoo it. A flexible marker pins the target area digitally — the same marker can recur across photos — so "the same lesion" is enforced by the software, not by ink or luck.
• Both the wide view and the close-up are tracked over time. A standardized macro series and a dermatoscopic marker series coexist and are followed longitudinally — the whole-skin context and the lesion detail, on one timeline.
• Protocols are yours. You encode your own views, regions, and cadence rather than adopting a vendor's fixed protocol, and free (unprotocolled) captures are supported for the real world.
• Multi-device and multilingual. A capture session spans the devices you already own — a dermatoscope-equipped phone for the close-up, another device for the macro views — synced into one patient record, in Portuguese, Spanish, or English, with LGPD-grade privacy by default.

The honest limits

Dermoscopy is powerful and it is operator-dependent: its accuracy rises with training and falls without it.^[2] The images and any derived scores document and help track visible findings; they are not, by themselves, a diagnosis — clinical interpretation belongs to a qualified professional, and automated analysis should support that judgment, not stand in for it. Imaginos is clinical documentation and analysis software — not a medical device — and does not make diagnoses. What it does is make dermoscopy's most valuable asset, the comparison over time, reliable: the same lesion, captured the same way, ready to read side by side.