Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Katharina Sies1, Julia K Winkler1, Christine Fink1

  • 1Abteilung Dermatologie, Universität Heidelberg.

Journal Der Deutschen Dermatologischen Gesellschaft = Journal of the German Society of Dermatology : JDDG
|June 17, 2021
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

[Multiple papules on the thighs in an adult].

Dermatologie (Heidelberg, Germany)·2026
Same author

Digital Dermatology.

Journal der Deutschen Dermatologischen Gesellschaft = Journal of the German Society of Dermatology : JDDG·2026
Same author

Multiclass convolutional neural network vs. 96 dermatologists in skin lesion diagnosis, an international study.

Journal der Deutschen Dermatologischen Gesellschaft = Journal of the German Society of Dermatology : JDDG·2026
Same author

The Endocytic Receptor DEC205 Is Expressed by Brain Endothelial Cells and Is Involved in Regulating the Blood-Brain Barrier.

Cells·2026
Same author

Performance of PEN-FAST and CEPH-FAST for Cephalosporin Allergy Delabeling.

Clinical and translational allergy·2026
Same author

Large-scale association study identifies lung cancer susceptibility copy number variants and their potential functional role in genetic instability.

medRxiv : the preprint server for health sciences·2026
Same journal

The different face of skin-rash induced by brentuximab vedotin.

Journal der Deutschen Dermatologischen Gesellschaft = Journal of the German Society of Dermatology : JDDG·2026
Same journal

Repositioning Non-Ablative Fractional Lasers - Molecular and Histological Insights from a scoping review.

Journal der Deutschen Dermatologischen Gesellschaft = Journal of the German Society of Dermatology : JDDG·2026
Same journal

Infusion-site reactions to foslevodopa/foscarbidopa in Parkinson's disease.

Journal der Deutschen Dermatologischen Gesellschaft = Journal of the German Society of Dermatology : JDDG·2026
Same journal

Tranexamic Acid-Soaked Intranasal Packing for Postoperative Bleeding After Nasal Interpolation Flap Repair.

Journal der Deutschen Dermatologischen Gesellschaft = Journal of the German Society of Dermatology : JDDG·2026
Same journal

Monoclonal gammopathy of dermatological significance: non-amyloid paraprotein deposits with light and heavy chains.

Journal der Deutschen Dermatologischen Gesellschaft = Journal of the German Society of Dermatology : JDDG·2026
Same journal

Papular epidermal nevus with "skyline" basal cell layer: Clinical clues from a three-member family.

Journal der Deutschen Dermatologischen Gesellschaft = Journal of the German Society of Dermatology : JDDG·2026
See all related articles

Convolutional neural networks (CNNs) show high accuracy in classifying skin lesions. However, large dark corner artifacts (DCAs) in dermoscopic images significantly reduce diagnostic specificity, highlighting system limitations for clinical use.

Area of Science:

  • Dermatology and Artificial Intelligence
  • Medical Image Analysis
  • Machine Learning in Healthcare

Context:

  • Artificial intelligence systems, specifically convolutional neural networks (CNNs), achieve dermatologist-level accuracy in skin lesion classification.
  • Understanding the limitations of these AI systems is crucial before widespread clinical adoption.
  • The impact of dark corner artifacts (DCAs) on AI diagnostic performance needs evaluation.

Purpose:

  • To investigate the influence of dark corner artifacts (DCAs) in dermoscopic images on the diagnostic performance of a commercially available CNN for skin lesion classification.
  • To assess how varying sizes of DCAs affect the sensitivity, specificity, and overall accuracy of the CNN.
  • To identify potential weaknesses in AI-powered dermatological diagnostic tools.

Summary:

Related Experiment Videos

  • A dataset of 233 skin lesion images was modified to include small, medium, and large DCAs.
  • A commercial CNN (Moleanalyzer-Pro®) analyzed 932 images for malignancy scores.
  • While small and medium DCAs had minimal impact, large DCAs significantly reduced specificity (87.9%) and led to higher malignancy scores, though overall ROC AUC remained high (0.962).

Impact:

  • CNN classification performance is not significantly affected by small or medium DCAs in dermoscopic images.
  • Large DCAs pose a challenge, leading to a significant decrease in specificity, indicating a limitation of the current AI system.
  • Clinicians must be aware of these technological limitations when submitting images for AI-assisted classification to ensure accurate diagnoses.