AI-assisted clinicians achieve 1.000 sensitivity in melanoma diagnosis but autonomous AI lacks clear clinical advantageArtificial Intelligence Shows Mixed Results in Detecting Melanoma

Frontiers in Medicine Published July 28, 2026 DOI ↗ Editorial oversight: Dr. Amelia Tan, PhD · Internal Medicine & Chronic Disease

AI-generated summary of the cited source, checked by automated accuracy review. How we work

Key Takeaway

Note that AI-assisted clinicians achieved 1.000 sensitivity, but autonomous AI shows no clear advantage over dermoscopy.

This meta-analysis evaluates the diagnostic performance of artificial intelligence (AI) for dermoscopic image analysis in patients with pigmented or melanoma-suspected skin lesions. The analysis synthesized 17 diagnostic arms from 10 studies, comparing AI alone, AI-assisted clinicians, and standard dermoscopy.

Key findings indicate that AI-assisted clinicians achieved a sensitivity of 1.000 and a specificity of 0.837. In contrast, autonomous AI showed overlapping performance with traditional dermoscopy; specifically, AI achieved slightly higher pooled specificity but lower sensitivity than the dermoscopy group. Dermoscopy sensitivity ranged from 0.418 to 0.966, while its specificity ranged from 0.293 to 0.975. Autonomous AI sensitivity ranged from 0.164 to 0.968 and specificity from 0.374 to 0.983.

The authors note substantial variability across different AI algorithms and clinical settings. Furthermore, evidence regarding the impact of AI on clinician performance is limited to a single diagnostic arm and is considered hypothesis-generating. Currently, there is no clear clinical advantage for autonomous AI over conventional dermoscopy, necessitating further research before practical recommendations can be made.

How this fits prior evidence

This meta-analysis addresses a gap in understanding how human-AI collaboration compares to autonomous systems. It builds upon previous evidence that AI diagnostic tools show high accuracy but limited real-world validation. While the current findings suggest higher sensitivity for AI-assisted clinicians, the authors note these results are hypothesis-generating rather than sufficient for immediate clinical implementation.

Researchers analyzed 17 different diagnostic tools across 10 studies to see how artificial intelligence (AI) performs when identifying melanoma. The study looked at both standalone AI systems and cases where AI assisted a human clinician in analyzing skin images.

The results showed that while AI can be used for diagnosis, it does not have a clear advantage over traditional dermoscopy. When AI worked alone, it showed varying levels of sensitivity and specificity. When AI was used to assist a human clinician, the sensitivity reached 1.000, but this specific data came from only one diagnostic arm.

Because there is so much variation between different AI programs and settings, these findings are currently considered hypothesis-generating. This means the results suggest interesting possibilities for the future but do not yet provide enough evidence to change how doctors treat patients today. More research is needed before AI can be recommended as a standard replacement for current methods.

What this means for you:

AI shows potential in skin cancer screening, but it does not currently outperform traditional clinical methods.

Common questions

Is AI better than a doctor at spotting skin cancer?

The study found no clear clinical advantage of autonomous AI over traditional dermoscopic assessment. While an AI-assisted clinician reached a sensitivity of 1.000, this data came from only one diagnostic arm and is considered hypothesis-generating rather than a reason to change current medical practices immediately.

How accurate is AI when used alone to find melanoma?

When used alone, the sensitivity of AI ranged from 0.164 to 0.968 and specificity ranged from 0.374 to 0.983. Because there is substantial variability across different algorithms, these results vary significantly depending on the specific system being used.

Can AI help doctors identify skin cancer more effectively?

The study compared AI-assisted clinicians to standard dermoscopy. While the data for assisted clinicians showed high sensitivity, the evidence is limited to a single diagnostic arm. More research is needed before these tools can be recommended as a standard part of clinical practice.

Study Details

Study typeMeta analysis

EvidenceLevel 1

PublishedJul 2026

View Original Abstract ↓

BackgroundAccurate risk stratification of pigmented skin lesions is essential for early melanoma detection and for reducing unnecessary excisions. Although artificial intelligence (AI) is increasingly applied to dermoscopic image analysis, its diagnostic performance compared to dermoscopy remains unclear.ObjectiveThis study aimed to compare the diagnostic performance of AI, dermoscopy, and AI-assisted clinicians for malignancy risk stratification of pigmented skin lesions.MethodsPubMed, Embase, Web of Science, and the Cochrane Library were systematically searched for studies evaluating AI, dermoscopy, or AI-assisted clinicians in diagnosing pigmented or melanoma-suspected skin lesions. Diagnostic performance metrics were calculated from extracted or reconstructed data, and study quality was assessed using the Quality Assessment of Diagnostic Accuracy Studies-2 tool (QUADAS-2) and QUADAS-comparative (QUADAS-C).ResultsA total of 2,571 records were identified, and 10 studies were included in the primary quantitative analysis, contributing 17 diagnostic arms. These included 10 dermoscopy arms, 6 AI-alone arms, and 1 AI-assisted clinician arm. In the dermoscopy group, sensitivity ranged from 0.418 to 0.966, and specificity ranged from 0.293 to 0.975. In the AI group, sensitivity ranged from 0.164 to 0.968, and specificity ranged from 0.374 to 0.983. AI-assisted clinicians showed a sensitivity of 1.000 and a specificity of 0.837 in the single available study. Overall, AI and dermoscopy showed overlapping diagnostic performance, although substantial variability was observed across AI algorithms and clinical settings. Deeks’ funnel plots showed no evidence of significant publication bias in either the AI group or the dermoscopy group.ConclusionAutonomous AI showed diagnostic performance broadly comparable to dermoscopy. Although AI achieved slightly higher pooled specificity, its sensitivity was lower, indicating no clear clinical advantage over conventional dermoscopic assessment. Evidence on AI-assisted clinicians was limited to a single diagnostic arm; therefore, this finding should be interpreted as hypothesis-generating rather than as evidence to support immediate clinical implementation. Further comparative studies using standardized AI-assisted workflows, comparable physician expertise, and patient-centered outcomes are required before practical recommendations can be made.Systematic review registrationRegistered with PROSPERO (CRD420261389834).

AI-assisted clinicians achieve 1.000 sensitivity in melanoma diagnosis but autonomous AI lacks clear clinical advantageArtificial Intelligence Shows Mixed Results in Detecting Melanoma

How this fits prior evidence

Common questions

Study Details

Evaluating the Efficacy of Traditional Chinese Medicine Decoctions for Rheumatoid Arthritis Management

Traditional Chinese medicine decoctions may improve rheumatoid arthritis symptoms

Clinical research that matters. Delivered to your inbox.

AI-assisted clinicians achieve 1.000 sensitivity in melanoma diagnosis but autonomous AI lacks clear clinical advantageArtificial Intelligence Shows Mixed Results in Detecting Melanoma

How this fits prior evidence

Common questions

More on Melanoma

Study Details

Evaluating the Efficacy of Traditional Chinese Medicine Decoctions for Rheumatoid Arthritis Management

Traditional Chinese medicine decoctions may improve rheumatoid arthritis symptoms

Clinical research that matters. Delivered to your inbox.

Related in Rheumatology

From Other Specialties