Dermoscopy — a clinical teaching reference

1. Introduction

What is dermoscopy?

Dermoscopy (also called dermatoscopy, epiluminescence microscopy [ELM] or skin-surface microscopy) is the in-vivo examination of skin lesions using a hand-held optical magnifier with a built-in light source. By eliminating surface reflection — either through cross-polarised light or an interface medium that optically couples the glass plate to the stratum corneum — dermoscopy reveals colours and structures in the epidermis, dermo-epidermal junction and superficial papillary dermis that are invisible to the unaided eye.

It is a non-invasive, real-time bedside technique. It does not replace histopathology but it materially shifts the pre-test probability of malignancy, supports the decision to biopsy or monitor, and is the foundation of digital lesion monitoring and AI-assisted triage.

Historical background

Johann Saphier first described "Dermatoskopie" in 1920 — oil-immersion examination of skin with a binocular microscope. Modern dermoscopy emerged in late-1980s Vienna with Pehamberger, Steiner and Wolff, who introduced systematic pattern analysis for pigmented lesions (1987). Subsequent international consensus conferences (Hamburg 1989, the Consensus Net Meeting on Dermoscopy 2003 and the International Dermoscopy Society [IDS] terminology consensus 2016) standardised the descriptive vocabulary and validated diagnostic algorithms used today.

Evidence base

• Trial Vestergaard meta-analysis (2008): dermoscopy by trained clinicians produced a ~9-fold improvement in the diagnostic odds ratio for melanoma versus naked-eye examination — sensitivity ~90% (vs 71%), specificity ~90% (vs 81%).
• Trial Cochrane review (Dinnes 2018): in-person dermoscopy by experienced clinicians had sensitivity 92% (95% CI 87–95%) and specificity 95% (95% CI 84–98%) for melanoma, exceeding visual inspection alone.
• Trial AI comparison (Tschandl, Lancet Oncol 2019): deep-learning algorithms reached comparable accuracy to human dermoscopists on image-based test sets — performance varied by lesion site and case mix.
• Consensus IDS terminology (Kittler 2016): standardised the metaphorical and descriptive language used across the discipline.
• Training effect: dermoscopy improves accuracy only with training. In untrained hands the sensitivity for melanoma can be lower than naked-eye examination — a recurring finding in primary-care studies.

Types of dermatoscope

• Non-polarised contact dermatoscope (traditional): requires a liquid interface medium (oil, alcohol gel, water or ultrasound gel) between the glass plate and the skin to eliminate surface reflection. Example: Heine Delta 20T (classic mode).
• Polarised contact dermatoscope: uses cross-polarised LED illumination; surface reflection is eliminated optically, so no interface fluid is required and skin can be examined dry. Modern default. Examples: DermLite DL4, 3Gen DL3N.
• Polarised non-contact dermatoscope: cross-polarised, examines without touching the skin. Useful for fragile, ulcerated, infected or painful lesions and for cross-infection control. Examples: Heine NC2, FotoFinder Handyscope (non-contact mode).
• Hybrid devices (polarised + non-polarised toggle): increasingly the standard, since different features are better seen in each mode. Examples: DermLite DL5, Heine Delta 30, FotoFinder Hands.
• Videodermoscopy / digital dermoscopy: higher magnification with image capture, total-body photography and AI-assisted analysis. Mainstream in UK pigmented-lesion clinics (e.g. FotoFinder, MoleMax, MoleScope).
• Smartphone-attached dermatoscopes: e.g. DermLite HÜD, FotoFinder Handyscope. Increasing primary-care use. Image quality is excellent in expert hands; the evidence base for primary-care triage with these devices is still developing and depends heavily on training.
• Reflectance confocal microscopy (RCM): an adjunct, not dermoscopy — in-vivo cellular-resolution (~3–5 µm) imaging used at specialist centres for difficult lentigo-maligna and amelanotic lesions.

Polarised vs non-polarised

Polarised and non-polarised dermoscopy show overlapping but distinct features. Many of the most diagnostically useful structures are visible in one mode but not the other — using both modes (or a hybrid device) is best practice.

Practical takeaway: default to polarised contact for most lesions; toggle to non-polarised contact (with fluid) when assessing a suspected seborrhoeic keratosis (milia, comedo openings) or when looking for blue-white veil in a pigmented lesion.

Interface medium

Required for non-polarised contact dermoscopy; optional for polarised contact (an interface fluid does slightly improve image quality even in polarised mode and is useful over scale and keratin).

• 70% isopropyl alcohol gel (hand sanitiser) — most practical, disinfects the skin and the lens, dries quickly, low viscosity. UK preferred default in clinics and primary care.
• Ultrasound gel — clear, biocompatible, preferred for mucosal, periocular, genital or ulcerated lesions where alcohol is contraindicated.
• Mineral oil — the traditional fluid; excellent optical coupling but messy and can damage some lens coatings.
• Water — adequate, universally available, evaporates quickly.
• Dedicated dermoscopy immersion fluids — available commercially; rarely necessary.

Avoid alcohol on broken skin, mucosa or near the eye. For nail-fold or subungual examination ultrasound gel works well.

2. Technique

Preparation

• Hand hygiene before and after each patient.
• Decontaminate the dermatoscope — alcohol wipe on the lens / glass plate between every patient. Replace any disposable lens covers between patients.
• Adequate ambient lighting; sit at the patient's level.
• Position the patient comfortably for the body site: supine for chest/abdomen, prone for back and posterior thigh, etc.
• Brief explanation and consent for examination and photography.

Choosing the mode

• Polarised contact — default for most pigmented and vascular lesions; quick, no fluid required.
• Non-polarised contact (with interface fluid) — switch to for suspected seborrhoeic keratoses (milia, comedo openings) and for blue-white veil. Use both modes routinely if the device has a toggle.
• Polarised non-contact — preferred for fragile or ulcerated lesions, suspected infectious lesions, mucosal sites and to minimise discomfort.

Holding the device

• Hold the dermatoscope like a pen.
• Brace your hand against the patient's skin so the field of view is steady — small involuntary movements blur fine structures.
• Use minimal contact pressure. Excess pressure blanches vessels and produces a false-negative for atypical vessels — an important and avoidable pitfall in amelanotic and pink lesions.
• For non-polarised mode, apply a small bead of interface fluid to the lesion or to the lens before contact.

Systematic scanning

• Scan the entire lesion in multiple orientations — asymmetry may be present along only one axis.
• Look from edge to centre and centre to edge, then rotate through 90° and repeat.
• Inspect immediately surrounding skin for context (background pattern, regression, satellite lesions).
• Apply the 'ugly duckling' sign — compare with the patient's other naevi. Most melanomas look different from a patient's typical naevus phenotype.

Total body skin examination (TBSE)

• Naked-eye TBSE first — identify the candidate lesion(s) by the ugly duckling sign or by patient report.
• Then dermoscopy of any candidate lesion.
• Document any lesion that is new, growing, changing colour or shape, ulcerating, bleeding or symptomatic.
• Examine acral (palms, soles), subungual, scalp and mucosal sites — these are commonly omitted from a quick examination but are exactly the sites at which acral lentiginous melanoma and lentigo maligna are missed.

Image documentation

• Capture a clinical photograph and a dermoscopic image at the same encounter.
• Include a measurement scale (or note the diameter in mm).
• Adequate focus and exposure; avoid over-exposure that washes out subtle pigment.
• Anonymise images per local information-governance policy. See the site's image consent SOP.

Sequential digital dermoscopy

• Capture and compare images at intervals (typically 3, 6 and 12 months) for patients with multiple atypical naevi or for borderline lesions where the threshold for excision has not yet been crossed.
• Any change in size, structure or colour over the interval is an indication for excision regardless of dermoscopy criteria, because slow-growing melanoma can lack dermoscopic atypia at presentation.
• Stable lesions over > 12 months are highly unlikely to be melanoma.

Cleaning and infection control

• 70% isopropyl alcohol wipe on the glass plate between every patient.
• Disposable plastic film or single-use covers for high-risk or open lesions; required for known blood-borne virus carriers.
• Avoid contact dermoscopy on broken skin, frank ulceration or mucosa without protection; use non-contact mode.

3. Assessment algorithms

No single algorithm is best for every clinician or every lesion. The hierarchy is: pattern analysis > validated scoring systems > naked-eye gestalt. Pattern analysis is the gold standard for the experienced dermoscopist; validated scoring systems are more useful for trainees and primary-care users.

3.1 The Two-Step Algorithm

The foundational diagnostic framework. Step 1 sorts melanocytic from non-melanocytic lesions; Step 2 applies pattern analysis (or a validated scoring system) only to lesions deemed melanocytic.

Step 1 — Is it melanocytic?

A lesion is melanocytic if any of the following are present:

• Pigment network (reticular pattern).
• Aggregated brown globules.
• Streaks (radial projections, pseudopods).
• Homogeneous blue pigmentation (blue naevus).
• Parallel patterns on acral skin.
• Pseudonetwork on facial skin.

If none of these are present, the lesion is non-melanocytic. Proceed to BCC, seborrhoeic keratosis, dermatofibroma, vascular lesion or other non-melanocytic pattern.

Caveat: some melanomas are featureless on dermoscopy (especially nodular and amelanotic melanoma). Step 1 is not a guarantee that an apparently non-melanocytic lesion is benign — always integrate with clinical context, change over time and patient risk.

Step 2 — Benign or malignant melanocytic?

Apply pattern analysis or a validated scoring system (see 3.2).

3.2 Pattern analysis and scoring systems

• Consensus Pattern analysis (Pehamberger / IDS). The gold standard for the trained dermoscopist — integrative recognition of global pattern, local features and colours. No single "score".
• Trial 3-point checklist (Soyer / Argenziano, 2004). Asymmetry, atypical network, blue-white structures. ≥ 2 features → biopsy. Sensitivity ~96% in primary-care use; designed for non-expert screening.
• Trial 7-point checklist (Argenziano, 1998; revised 2011). Three major features (atypical pigment network, blue-white veil, atypical vascular pattern — 2 points each) plus four minor features (regression, irregular streaks, irregular dots/globules, irregular blotches — 1 point each). Total ≥ 3 → biopsy.
• Trial ABCD rule of dermoscopy (Stolz, 1994). Asymmetry, Border, Colour, Differential structures → total dermoscopy score (TDS): < 4.75 benign; 4.75–5.45 borderline / suspicious; > 5.45 highly suspicious for melanoma.
• Trial Menzies method (1996). Symmetry pattern absent + at least one of nine positive features (blue-white veil, multiple brown dots, pseudopods, radial streaming, scar-like depigmentation, peripheral black dots/globules, multiple colours, multiple blue/grey dots, broadened network).
• Consensus CASH algorithm (Henning, 2007). Colour, Architecture, Symmetry, Homogeneity.
• Consensus Chaos and Clues (Rosendahl, 2012). A simplified two-step approach used in Australian primary care — chaos (asymmetry of structure or colour) plus at least one clue (e.g. grey/blue, eccentric structureless area, atypical network, polymorphous vessels) → biopsy.
• Sequential digital dermoscopy — not an algorithm but a key adjunct for monitoring borderline lesions; see 2.7.

3.3 By lesion type

Pigmented melanocytic lesions

Apply the Two-Step Algorithm and pattern analysis or a scoring system. Consult the pigmented structures table for individual features. Key concerning features are atypical network, blue-white veil, atypical vessels and regression structures — the major criteria of the 7-point checklist.

Basal cell carcinoma

Diagnosis rests on the presence of one or more BCC criteria together with the absence of melanocytic criteria:

• Arborising vessels — large-calibre, branching tree-like vessels in sharp focus on the surface. Highly characteristic for BCC in the right clinical context.
• Blue-grey ovoid nests — large blue-grey confluent areas (pigmented BCC).
• Blue-grey globules / multiple grey-blue dots — smaller pigment aggregates.
• Leaf-like (maple-leaf) areas — discrete brown-grey peripheral bulbous extensions.
• Spoke-wheel structures — radial projections meeting at a darker hub.
• Concentric structures — a newer BCC criterion.
• Ulceration — central erosion, often with adherent serum.
• Shiny white (chrysalis) structures — on polarised dermoscopy.
• Absence of pigment network — an important negative criterion.

≥ 1 BCC criterion + absence of melanocytic criteria suggests BCC. Confirm with biopsy. See the BCC monograph.

cSCC and Bowen's disease

Bowen's disease (in-situ SCC)

• Glomerular (coiled) vessels — small tightly-coiled vessels resembling renal glomeruli. Characteristic.
• Dotted vessels in clusters.
• Yellow scales or surface keratin.
• Pinkish background.

Invasive cSCC

• Polymorphous vessels (mixed dotted, hairpin and linear-irregular).
• White circles around hair follicles ('targetoid' appearance).
• Keratin / scale on the surface; central ulceration.
• White structureless areas (keratinisation).

See the cSCC monograph and the Bowen's disease monograph.

Seborrhoeic keratoses (a common mimic)

• Comedo-like openings (black or brown plugs).
• Milia-like cysts (round white-yellow structures — better seen on non-polarised mode).
• 'Fingerprint' or 'cerebriform' surface pattern.
• Sharp demarcation, 'stuck-on' appearance.
• Hairpin vessels.
• Pseudonetwork-like reticular structures.

Mistaking pigmented seborrhoeic keratosis for melanoma is common. Conversely, melanomas can mimic seborrhoeic keratoses (the 'false-negative trap'). When uncertain, biopsy.

Dermatofibroma

• Central white scar-like patch with peripheral pigment network — the classical 'doughnut' appearance.
• Symmetrical brown structureless area.
• Dotted vessels within the central area.

Vascular and angiomatous lesions

• Cherry angioma / haemangioma: red or red-blue lacunes (well-demarcated rounded vascular spaces) with white intervening septa.
• Thrombosed angioma: dark blue-black lacunes; can mimic nodular melanoma — the 'red lagoon vs dark melanoma' problem. Look for surrounding red lacunes.
• Angiokeratoma: dark lacunes plus white-yellow keratotic surface scale.
• Pyogenic granuloma: red-pink homogeneous structureless area with a white collarette; can mimic amelanotic melanoma — biopsy in doubt.

3.4 By anatomical site

Face and lentigo maligna

Facial skin has a thin epidermis and prominent appendages (hair follicles, sebaceous glands). Dermoscopic structures therefore differ from those on the trunk and limbs:

• Pseudonetwork — pigmented background interrupted by non-pigmented holes that correspond to follicular openings (not true rete-ridge network).
• Annular-granular pattern — fine grey dots around follicular openings. Concerning for lentigo maligna.
• Asymmetric pigmented follicular openings — concerning for lentigo maligna.
• Rhomboid (or zigzag) structures — coalescing pigmented lines around follicles. Concerning for lentigo maligna.
• Grey-blue pseudonetwork — concerning for lentigo maligna / lentigo maligna melanoma.
• Obliterated follicles — loss of follicular pattern — concerning, suggests dermal invasion.
• Solar lentigo: homogeneous light brown pseudonetwork, sharply defined, with a 'moth-eaten' or 'jelly sign' edge — benign.
• Pigmented actinic keratosis: pseudonetwork that can mimic LM; clues are surface scale and follicular hyperkeratosis. The four-point LM algorithm of Pralong / Tschandl identifies LM with high specificity but inter-rater agreement is moderate — biopsy if in doubt.

See the lentigo maligna monograph.

Palms, soles and acral skin

Acral skin has parallel ridges and furrows (dermatoglyphics) which dictate the patterns seen. Acral lentiginous melanoma is proportionally commoner in Fitzpatrick IV–VI skin and is a recognised equity gap in UK practice — have a low threshold for biopsy on plantar / palmar / digital lesions.

Benign patterns

• Parallel furrow pattern — pigment in the furrows (the depressed lines). Most common benign pattern.
• Lattice-like pattern — variant of parallel furrow with crossing lines (commonly on the arch).
• Fibrillar pattern — fine parallel lines obliquely crossing furrows. Common on weight-bearing soles.

Concerning for acral melanoma

• Trial Parallel ridge pattern — pigment on the ridges (the raised lines, where eccrine ducts open). A high-specificity sign in trained hands; the original Saida series reported sensitivity in the mid-80% range and specificity ~99% for early acral melanoma, but performance varies with training and lesion mix.
• Disorganised, asymmetric pigmentation; multiple colours.
• Ulceration, blue-white veil or atypical vessels in an acral lesion.

Subungual lesions

• Longitudinal melanonychia — assess width (≥ 3 mm in adults is a relative red flag per the ABCDEF rule: Age / race, Brown-black band ≥ 3 mm / Border irregularity, Change, Digit involvement (thumb / index / hallux), Extension of pigment to periungual skin (Hutchinson's sign), Family / personal history), colour homogeneity and parallel-line regularity.
• Triangular shape (wider proximally than distally) is concerning.
• Disrupted parallel lines (lines that vary in colour, width or spacing) vs uniform parallel lines.
• Periungual pigmentation (pseudo-Hutchinson in some benign cases; true Hutchinson in subungual melanoma).
• Refer for nail-matrix biopsy if any feature is concerning, particularly in skin of colour where acral lentiginous melanoma is more common.

Mucosal lesions

• Use polarised non-contact or non-polarised contact with ultrasound gel (not alcohol).
• Benign mucosal melanosis typically shows structureless light-brown pigmentation; melanotic macules show parallel or fish-scale patterns.
• Concerning features for mucosal melanoma: multiple colours including blue / grey / white, structureless areas, asymmetry. Threshold for biopsy is low because dermoscopic criteria are less validated than for skin.

Scalp and hair-bearing sites

• Scalp melanomas are commonly amelanotic / hypomelanotic and are frequently late-diagnosed — hair obscures the lesion and there are large empty follicular openings. A high index of suspicion is required for any solitary, growing, ulcerating or bleeding scalp lesion in an older patient.
• Pigmented scalp lesions in adults: think melanoma (uncommon site, late presentation) and pigmented BCC.

3.5 Special considerations

Pink / amelanotic lesions

• Vessels carry the diagnostic weight when pigment is absent. Polymorphous vessels, milky-red areas and shiny white structures are concerning for amelanotic melanoma.
• Pizzichetta (2004) emphasised: any growing pink nodule in an adult should be biopsied; nodular amelanotic melanoma is a high-risk presentation that is easily missed.

Nodular melanoma

• Often featureless on dermoscopy — classical criteria may be absent.
• Clues: rapid growth, blue-black colour, peripheral atypical vessels, ulceration, asymmetry of pigmentation.
• The EFG rule (Elevated, Firm, Growing for > 1 month) is more sensitive than dermoscopy for nodular melanoma — integrate with clinical examination.

Skin of colour

• Acral lentiginous melanoma is proportionally commoner in Fitzpatrick IV–VI skin; index of suspicion for plantar, palmar and subungual lesions must remain high.
• Dermoscopy of pigmented lesions in skin of colour shows the same structures but on a darker baseline; experience and a calibrated white-balance image matter.
• Subungual streaks are common and frequently benign in Fitzpatrick V–VI; abrupt change, disrupted parallel lines, periungual extension or width > 3 mm should still prompt review and likely biopsy.
• Vascular structures and shiny white features are equally diagnostic.

4. Appendix — dermoscopic features

Categorised reference tables for individual dermoscopic features. Terminology follows the International Dermoscopy Society consensus (Kittler 2016).

4.1 Pigmented structures

4.2 Vascular patterns

4.3 Surface and morphological features

4.4 Site-specific patterns

4.5 BCC-specific structures