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

A multiple classifier system for early melanoma diagnosis.

Andrea Sboner1, Claudio Eccher, Enrico Blanzieri

  • 1ITC-irst, Centre for Scientific and Technological Research, Via Sommarive 18, Povo, Trento 38050, Italy. sboner@itc.it

Artificial Intelligence in Medicine
|December 11, 2002
PubMed
Summary

Combining multiple classifiers significantly improves melanoma diagnosis accuracy. This approach matches dermatologist performance, offering a promising tool for early skin cancer detection.

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

Patient-derived organoids across cancers reveal conserved tumor heterogeneity and actionable therapeutic vulnerabilities.

Science advances·2026
Same author

Enhancing the quality and trustworthiness of large language model-generated summaries of clinical oncology literature.

JAMIA open·2026
Same author

Tumor immune microenvironment reconstitution in patient-derived organoids enables therapy modeling for NSCLC.

Cell reports methods·2026
Same author

Clinical and Mechanistic Evidence for Comano Thermal Water: A Narrative Review.

International journal of molecular sciences·2026
Same author

DEMETRA: An ACT-Based Virtual Coach to Support Healthier Lifestyles in Overweight Pregnant Women-Protocol for a Feasibility Pilot Study.

International journal of environmental research and public health·2026
Same author

Whole genome sequencing of locally advanced and metastatic breast carcinoma unravels relevant molecular signatures and novel events.

Pathology, research and practice·2026

Area of Science:

  • Dermatology
  • Computer Science
  • Medical Imaging

Background:

  • Melanoma is a dangerous skin cancer where early diagnosis is crucial for successful treatment.
  • Dermatologists achieve approximately 80% sensitivity and specificity through visual inspection.
  • Existing computerized diagnostic systems use various classification algorithms with varying success.

Purpose of the Study:

  • To develop a novel approach combining multiple classifiers to enhance diagnostic performance in melanoma detection.
  • To improve upon the diagnostic accuracy of single classification algorithms.

Main Methods:

  • Utilized three classifiers: linear discriminant analysis (LDA), k-nearest neighbour (k-NN), and a decision tree.
  • Extracted 38 geometric and colorimetric features from digital skin lesion images.

Related Experiment Videos

  • Generated multiple classifier systems by combining single classifier outputs using voting schemata and evaluated on 152 images.

    • Main Results:

    • Three-classifier systems showed significantly higher performance than 1- and 2-classifier systems (P<0.0005 and P<0.001).
    • No significant performance difference was observed between 1- and 2-classifier groups (P=0.352).
    • Multiple classifier groups demonstrated comparable performance to a group of eight dermatologists, outperforming single classifiers (P<0.020).

    Conclusions:

    • A suitable combination of different classifiers can significantly improve the performance of automated diagnostic systems for melanoma.
    • This multi-classifier approach achieves diagnostic accuracy comparable to expert dermatologists.
    • The findings suggest a promising direction for developing advanced computer-aided diagnostic tools for skin cancer.