CoChat
Sign Up
Cited by 285AI in cancer detectionCutaneous Melanoma Detection and ManagementArtificial Intelligence in Healthcare and Education

Deep learning outperformed 11 pathologists in the classification of histopathological melanoma images

Open access
Hekler, Achim, Utikal, Jochen S., Enk, Alexander H., Solass, Wiebke, Schmitt, Max, Klode, Joachim, Schadendorf, Dirk, Sondermann, Wiebke +7 more·2019·European Journal of Cancer
DOI 10.1016/j.ejca.2019.06.012
Open this paper in CoChat — free account

Abstract

BACKGROUND: The diagnosis of most cancers is made by a board-certified pathologist based on a tissue biopsy under the microscope. Recent research reveals a high discordance between individual pathologists. For melanoma, the literature reports on 25-26% of discordance for classifying a benign nevus versus malignant melanoma. A recent study indicated the potential of deep learning to lower these discordances. However, the performance of deep learning in classifying histopathologic melanoma images was never compared directly to human experts. The aim of this study is to perform such a first direct comparison. METHODS: A total of 695 lesions were classified by an expert histopathologist in accordance with current guidelines (350 nevi/345 melanoma). Only the haematoxylin & eosin (H&E) slides of these lesions were digitalised via a slide scanner and then randomly cropped. A total of 595 of the resulting images were used to train a convolutional neural network (CNN). The additional 100 H&E image sections were used to test the results of the CNN in comparison to 11 histopathologists. Three combined McNemar tests comparing the results of the CNNs test runs in terms of sensitivity, specificity and accuracy were predefined to test for significance (p < 0.05). FINDINGS: The CNN achieved a mean sensitivity/specificity/accuracy of 76%/60%/68% over 11 test runs. In comparison, the 11 pathologists achieved a mean sensitivity/specificity/accuracy of 51.8%/66.5%/59.2%. Thus, the CNN was significantly (p = 0.016) superior in classifying the cropped images. INTERPRETATION: With limited image information available, a CNN was able to outperform 11 histopathologists in the classification of histopathological melanoma images and thus shows promise to assist human melanoma diagnoses.

Integrity signals

Registry

Crossref · 10.1016/j.ejca.2019.06.012

Cites

23 of 23 references verified0 citation errors · 0 not checkable

Statistics

Recomputing p-values needs the paper’s full text.

Claims the paper makes

We couldn’t retrieve this paper’s full text, so its claims and self-check weren’t run. The registry and citation checks still apply.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • Hekler et al., 2019, using a CNN model, achieved a mean sensitivity of 76%, specificity of 60%, and accuracy of 68% in classification of melanoma histologic images, while pathologists achieved a mean sensitivity of 51.8%, specificity of 66.5%, and accuracy of 59.2%.

    “Hekler et al., 2019 [17], using a CNN model, achieved a mean sensitivity of 76%, specificity of 60%, and accuracy of 68% in classification of melanoma histologic images. In comparison, the pathologists achieved a mean sensitivity of 51.8%, specificity of 66.5%, and accuracy of 59.2%.”

    From Artificial Intelligence in Dermatopathology: An Update and Review of the Current Literature (2026) · cites this paper as [17]

  • Deep learning algorithms have shown strong performance in the histopathologic evaluation of melanocytic lesions, often comparable to or exceeding human experts.

    “Deep learning algorithms have shown strong performance in the histopathologic evaluation of melanocytic lesions, often comparable to or exceeding human experts”

    From Artificial Intelligence in Dermatopathology: An Update and Review of the Current Literature (2026) · cites this paper as [17]

Cites

Flagged first, then by citations

Cited by

Sorted by citations

Related

Authors
Hekler, A., Utikal, J. S., Enk, A. H., Solass, W., Schmitt, M., Klode, J., Schadendorf, D., Sondermann, W., Franklin, C., Bestvater, F., Flaig, M. J., Krahl, D., von Kalle, C., Fröhling, S., & Brinker, T. J.
Affiliations
  • German Cancer Research Center (DE)
  • University Hospital Heidelberg (DE)
  • University of Tübingen (DE)
  • Essen University Hospital (DE)
  • University Hospital Cologne (DE)
  • LMU Klinikum (DE)
Corresponding author
Brinker, Titus J.
ORCID
Venue
European Journal of Cancer
Publisher
Elsevier BV
Published
2019-09
Volume
118
Pages
91-96
ISSN
0959-8049 (ISSN-L)
DOI
10.1016/j.ejca.2019.06.012
OpenAlex
W2963258365
Cited by
285
References
23
Citations by year
2026: 21 · 2025: 34 · 2024: 40 · 2023: 41 · 2022: 52 · 2021: 44 · 2020: 42 · 2019: 10
FWCI
17.07
Type
journal article
Language
en
Open access
Licence
Locations
  • European Journal of Cancer (publishedVersion, cc-by-nc-nd)
  • PubMed (publishedVersion)
  • Kölner Universitäts PublikationsServer (Universität zu Köln) (publishedVersion)
Topics
  • Physical Sciences › Computer Science › Artificial Intelligence › AI in cancer detection
  • Health Sciences › Medicine › Oncology › Cutaneous Melanoma Detection and Management
  • Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
Keywords
melanoma, histopathology, convolutional neural networks, deep learning, pathologist agreement, diagnostic discordance, benign nevus, H&E slides, computer-aided diagnosis, skin cancer diagnosis
MeSH terms
Pathologists, Deep Learning, Biopsy, Diagnosis, Computer-Assisted, Diagnosis, Differential, Humans, Image Interpretation, Computer-Assisted, Melanoma, Microscopy, Nevus, Predictive Value of Tests, Skin Neoplasms, Reproducibility of Results, Observer Variation

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review