Dermoscopy and image-based diagnosis

Scope of this leaf

This module is the application / image-diagnosis companion to dermoscopy-principles. Focus: decision algorithms, pattern dictionaries, special sites, monitoring vs excision thresholds, and AI/teledermoscopy limits for board exams (FRCDerm, ABD, FACD, MRCP SCE, NEET-PG/IADVL). [1]

Instrumentation quick recap

• Non-polarised immersion: liquid interface required; excellent for superficial pigment and keratin plugs.[1]
• Polarised: no fluid; better for vessels and shiny white lines (chrysalis).[1]
• Hybrid devices toggle modes — use both when a lesion is equivocal.

The diagnostic algorithm

Step 1 — Melanocytic or not?

Melanocytic if any classic criterion is present: pigment network, aggregated globules, streaks/pseudopods, homogeneous blue pigmentation, or parallel pattern (acral/mucosa).[1][2] If none → non-melanocytic pathway (BCC, SK, vascular, keratinocyte neoplasia, dermatofibroma, etc.).

Step 2 — Within melanocytic lesions

Use pattern analysis (asymmetry of colours/structures) and/or checklists: [1]

IDS web-based work supports that trained criteria can be applied with measurable validity and reliability — training matters.[2]

Chaos and clues (routine practice)

1. Is there chaos (asymmetric colour or structure)?
2. If yes, seek clues (e.g. eccentric structureless areas, thick reticular lines, grey/blue structures, peripheral black dots/globules, pseudopods, white lines, polymorphous vessels).
3. Chaos + clue(s) → biopsy/excise rather than long-term observation.[3][4]

Organised lesions without clues may be monitored or left — unless history (change, symptoms) or special-site rules intervene. [1]

Vascular and structure dictionary

Vascular decoding and keratinocyte progression features (AK → SCC) are high-yield for image stations.[1][9]

Special sites

Face (lentigo maligna pathway)

Sun-damaged facial skin uses pseudonetwork architecture. Concerning: asymmetric pigmented follicular openings, rhomboidal structures, annular-granular pattern, progressive obliteration of follicular openings.[7] Partial mapping biopsies may be guided by the most dermoscopically atypical sectors of a large patch.

Acral

Parallel ridge pattern (pigment on ridges that bear eccrine openings) is a melanoma red flag; parallel furrow, lattice, and fibrillar patterns favour acral naevi when classic.[6][7] Do not apply truncal network rules naively to palms and soles.

Nail

Assess band width, colour homogeneity, border regularity, and periungual pigment (micro-Hutchinson). Single-digit progressive melanonychia in adults needs a low threshold for specialist matrix assessment.[7]

Dermoscopy–histology correlation

Understanding correlates prevents magical thinking: [1]

• Pigment network ↔ melanin along elongated rete ridges.[1]
• Globules ↔ melanocytic nests.
• Blue structures ↔ deeper dermal melanin (Tyndall).
• Blue-white veil ↔ compact orthokeratosis over melanin-rich dermis.
• Shiny white lines ↔ collagen remodel under polarised light.

Correlation also explains why dermoscopy improves biopsy targeting and can reduce excision of stereotypical benign SK/BCC patterns in expert hands — without replacing histology for suspicious lesions.[1]

Beyond tumours: inflammoscopy, trichoscopy, entomodermoscopy

• Inflammoscopy: regularly distributed dotted vessels on a red background suggest psoriasis; Wickham striae favour lichen planus; patterns refine differentials among common inflammatory dermatoses.[8]
• Trichoscopy: yellow dots, black dots, broken hairs and exclamation-mark hairs support alopecia areata; white dots / loss of follicular openings favour scarring alopecias — the dermatoscope is now standard in hair clinics.[12]
• Entomodermoscopy: classic scabies “delta wing / jet with contrail” sign speeds bedside confirmation (details also in scabies atlas leaf).

Digital monitoring, teledermoscopy, and AI

Short-term digital dermoscopy monitoring

Appropriate for flat, feature-poor, low-suspicion melanocytic lesions when immediate excision is not mandatory. Combine static morphology with dynamic change calculators/features; clear growth, new colours, or new melanoma clues end monitoring.[11] Exclude nodular, ulcerated, amelanotic red-flag, and high-anxiety non-compliant contexts.

Teledermoscopy

Requires sharp focus, correct mode (polarised/immersion), macroscopic context photos, and clinical history. Garbage-in images produce garbage-out triage. [1]

AI

Convolutional networks have reached dermatologist-level classification performance on curated image tasks (landmark Esteva 2017 Nature work), but real-world deployment faces dataset shift, skin-of-colour under-representation, rare entities, and medicolegal accountability.[10] Exam-safe stance: AI may assist; clinicopathologic correlation decides.

Decision algorithm: observe, monitor, or biopsy

1. Clinical context (history of change, immunosuppression, prior melanoma).
2. Macroscopic ABCDE / EFG for nodules.
3. Dermoscopy two-step + site rules.
4. Excise/biopsy now if melanoma-specific features, parallel ridge, LM facial clues, micro-Hutchinson pathway, or polymorphous vessels in a concerning lesion.
5. Short-term digital monitor only if flat, low-suspicion, and follow-up assured.[11]
6. Histology remains gold standard for definitive diagnosis of excised tissue.

Preferred biopsy when melanoma is realistic: full-thickness excisional biopsy with narrow margins (or carefully planned partial sampling of large LM) — do not rely on superficial shave that understages Breslow. [1]

Pitfalls

• SK-like verrucous melanoma and collision tumours.
• Recurrent naevus (pseudomelanoma) after incomplete shave — history is everything.
• Over-reliance on a single criterion (e.g. any dotted vessels).
• Monitoring a nodular pink lesion because “dermoscopy was inconclusive.”
• Trainee overconfidence without structured pattern training.[2]

Regional notes

Core pattern language is global (IDS terminology). Resource-limited settings still gain from handheld dermoscopy for triage; AI smartphone tools need the same red-flag discipline as clinic devices.[10]

Clinical pearl

Red flags

Exam application bank (NEET-PG / INICET)

One-line answer

Board-level application module for dermoscopy as image-based diagnosis: two-step algorithm, chaos-and-clues pattern analysis, vascular morphology dictionary, non-melanocytic tumour patterns, special-site rules (face, acral, nail), dermoscopy–histology correlation, inflammoscopy and trichoscopy snapshots, digital monitoring thresholds, teledermoscopy quality, AI-assisted classification limits, and a red-flag biopsy algorithm. Complements the principles leaf with decision-focused exam content.

Worked stems (answer without another resource)

Stem 1 — Classic presentation. Map symptoms to mechanism; name the first investigation and first treatment step with dose/route if drug therapy is standard. [1]

Stem 2 — Unstable / complicated. List red flags that force immediate resuscitation, theatre, ICU, antidote, or reperfusion — and what you do in the first 15 minutes. [1]

Stem 3 — Atypical group. Elderly, pregnancy, child, or immunocompromised: how presentation and thresholds change. [1]

Stem 4 — Differential trap. Name the three closest mimics and one discriminator for each. [1]

Stem 5 — Disposition. Who goes home with safety-netting, who is admitted, who needs HDU/ICU/theatre, and what follow-up is mandatory. [1]

Rapid viva checklist

1. Definition + classification
2. Pathophysiology chain
3. Bedside signs / criteria
4. Score with exact components (if any)
5. Emergency bundle
6. Definitive therapy with doses
7. Complications of disease and of treatment
8. Special populations
9. Guideline/trial name if classic
10. Three exam traps

Coverage self-check

If you cannot answer any stem above from this page alone, re-read the matching section — the page is intended to be self-sufficient for final-prof and NEET-PG/INICET questions on Dermoscopy and image-based diagnosis.

Expanded exam teaching (depth pass)

Clinical reasoning

For Dermoscopy and image-based diagnosis, examiners test whether you can prioritise life threats, choose the right first test, and give specific therapy (agent, dose, route, timing). Generic phrases without numbers score poorly.

Mechanism → feature map

Build a short chain: cause → pathophysiologic intermediate → clinical feature → complication. Every major symptom in the classic vignette should sit on that chain.

Investigation strategy

• Bedside/first-line tests that change immediate management
• Confirmatory or staging tests
• What a normal result does not exclude
• When not to delay treatment for imaging (unstable patient)

Management ladder

1. Resuscitation / ABC / sepsis or haemorrhage bundle as relevant
2. Specific antidote / procedure / antimicrobial / reperfusion / surgery
3. Supportive care and monitoring targets
4. Definitive long-term therapy and secondary prevention
5. Disposition and safety-net advice

Special populations

Always prepare one line each for children, pregnancy, elderly, renal/hepatic impairment, and immunocompromised patients when the topic allows.

Pitfalls that fail candidates

• Treating the number not the patient
• Missing pregnancy status when relevant
• Imaging before stabilisation
• Wrong empiric cover or wrong antidote timing
• Incomplete counselling on recurrence, adherence, or red-flag return

Board-level application module for dermoscopy as image-based diagnosis: two-step algorithm, chaos-and-clues pattern analysis, vascular morphology dictionary, non-melanocytic tumour patterns, special-site rules (face, acral, nail), dermoscopy–histology correlation, inflammoscopy and trichoscopy snapshots, digital monitoring thresholds, teledermoscopy quality, AI-assisted classification limits, and a red-flag biopsy algorithm. Complements the principles leaf with decision-focused exam content. [1]

Structured revision sheet

Must-know numbers and names

List every score, size threshold, dose, and time window from this topic on a blank page from memory, then check against the sections above.

Three classic MCQ angles

1. Most likely diagnosis given a vignette
2. Next best step in management
3. Most appropriate investigation

Three classic SAQ angles

1. Pathophysiology in five steps
2. Management algorithm with doses
3. Complications and prevention

Clinical station flow

Greet → focused history → targeted exam → investigations → explain diagnosis → emergency care → definitive plan → safety-net / follow-up → answer examiner questions on mechanism and pitfalls.

References

1. [1]Yélamos O, Braun RP, Liopyris K, Wolner ZJ, et al. Dermoscopy and dermatopathology correlates of cutaneous neoplasms J Am Acad Dermatol, 2019.PMID 30321581
2. [2]Carrera C, Marchetti MA, Dusza SW, Argenziano G, et al. Validity and Reliability of Dermoscopic Criteria Used to Differentiate Nevi From Melanoma: A Web-Based International Dermoscopy Society Study JAMA Dermatol, 2016.PMID 27074267
3. [3]Rosendahl C, Cameron A, McColl I, Wilkinson D. Dermatoscopy in routine practice - 'chaos and clues' Aust Fam Physician, 2012.PMID 22762066
4. [4]Ramji R, Valdes-Gonzalez G, Oakley A, Rademaker M. Dermoscopic 'Chaos and Clues' in the diagnosis of melanoma in situ Australas J Dermatol, 2018.PMID 29094749
5. [5]Nachbar F, Stolz W, Merkle T, Cognetta AB, et al. The ABCD rule of dermatoscopy. High prospective value in the diagnosis of doubtful melanocytic skin lesions J Am Acad Dermatol, 1994.PMID 8157780
6. [6]Saida T, Koga H, Uhara H. Key points in dermoscopic differentiation between early acral melanoma and acral nevus J Dermatol, 2011.PMID 21175752
7. [7]Thomas L, Phan A, Pralong P, Poulalhon N, et al. Special locations dermoscopy: facial, acral, and nail Dermatol Clin, 2013.PMID 24075549
8. [8]Sgouros D, Apalla Z, Ioannides D, Katoulis A, et al. Dermoscopy of Common Inflammatory Disorders Dermatol Clin, 2018.PMID 30201145
9. [9]Álvarez-Salafranca M, Zaballos P. [Translated article] Dermoscopy of Squamous Cell Carcinoma: From Actinic Keratosis to Invasive Forms Actas Dermosifiliogr, 2024.PMID 39102978
10. [10]Esteva A, Kuprel B, Novoa RA, Ko J, et al. Dermatologist-level classification of skin cancer with deep neural networks Nature, 2017.PMID 28117445
11. [11]Zenone M, Zocchi L, Moccia C, Passerini SG, et al. Digital dermoscopy monitoring of melanocytic lesions: Two novel calculators combining static and dynamic features to identify melanoma J Eur Acad Dermatol Venereol, 2022.PMID 34862986
12. [12]Pirmez R. The dermatoscope in the hair clinic: Trichoscopy of scarring and nonscarring alopecia J Am Acad Dermatol, 2023.PMID 37591567