Skin lesions of face and scalp – Classification by a market-approved convolutional neural network in comparison with 64 dermatologists

Haenssle, Holger A.; Winkler, Julia; Fink, Christine; Toberer, Ferdinand; Enk, Alexander; Stolz, Wilhelm; Deinlein, Teresa; Hofmann‐Wellenhof, Rainer; Kittler, Harald; Tschandl, Philipp; Rosendahl, Cliff; Lallas, Aimilios; Blum, Andreas; Abassi, Mohamed Souhayel; Thomas, Luc; Tromme, Isabelle; Rosenberger, Albert; Bachelerie, M.; Bajaj, Sonali; Balcere, Alise; Baricault, Sophie; Barthaux, Clément; Beckenbauer, Yvonne; Bertlich, Ines; Bouthenet, Marie-France; Brassat, Sophie; Buck, Philipp Marcel; Buder‐Bakhaya, Kristina; Cappelletti, Maria-Letizia; Chabbert, Cécile; Labarthe, Julie De; DeCoster, Eveline; Dobler, Michèle; Dumon, Daphnée; Emmert, Steffen; Gachon-Buffet, Julie; Gusarov, Mikhail; Hartmann, Fritz; Hartmann, Julia; Herrmann, Anke; Hoorens, Isabelle; Hulstaert, Eva; Karls, Raimonds; Kolonte, Andreea; Kromer, Christian; Vasseux, Céline Le Blanc; Levy-Roy, Annabelle; Majenka, Pawel; Marc, Marine; Bourret, Veronique Martin; Michelet-Brunacci, Nadège; Mitteldorf, Christina; Paroissien, Jean; Picard, Camille; Plise, Diana; Reymann, V.; Ribeaudeau, Fabrice; Richez, Pauline; Plaine, Hélène Roche; Salik, Dany; Sattler, Elke; Schäfer, Sarah; Schneiderbauer, Roland; Secchi, Thierry; Talour, Karen; Trennheuser, Lukas; Wald, Alexander; Wölbing, Priscila; Zukervar, P.

doi:10.1016/j.ejca.2020.11.034

Cited by 31 publications

(20 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Moreover, Haenssle et al [24] proposed a reader study that focused exclusively on suspicious lesions of the face and scalp. In level II of that study, the CNN significantly outperformed 64 human experts in terms of management decision (at dermatologists' specificity of 69.4%: 96.2% vs. 84.2% sensitivity, p < 0.001).…”

Section: Automated Skin Cancer Classification Of Dermoscopic Imagesmentioning

confidence: 99%

“…Three dermoscopic approaches expanded on the binary perspective (e.g. MM vs. melanocytic nevus, benign vs. malignant) presented by Brinker et al [16,17], Yu et al [19], Marchetti et al [20,21] and Haenssle et al [4,18,24], by carrying out multiclass classification tasks which covered more fine-grained diagnoses (see Table 1) [7,22,23]. Supplementary Table 4 outlines similarities and differences of these multiclass approaches with regard to individual training and testing procedures.…”

Section: Automated Skin Cancer Classification Of Dermoscopic Imagesmentioning

confidence: 99%

“…On the other hand, the images of the ISIC database (used as a test set for 6 out of 11 dermoscopic reader studies) mainly encompassed light-skinned skin lesions from patients in Europe, Australia and the United States, whereas Asian and dark-skinned populations were underrepresented. Yu et al [19], Haenssle et al [24] and Tschandl et al [7] proved the potential of CNN-based classification for special anatomic sites such as the face and scalp [24] or acral MM on the hands and feet [19], as well as rare subtypes such as amelanotic MM [7]. However, other special anatomic sites (e.g.…”

Section: Test Set Characteristicsmentioning

confidence: 99%

“…The decision-making basis of 14 of the 19 (see Tables 1e3) included reader studies was limited to a single image of the suspicious skin lesion. Haenssle et al [4,18,24] showed that dermatologists performed somewhat better, when provided with additional close-up images and patient information such as age, sex or lesion location. The authors highlighted the value of clinical data in addition to visual data.…”

Section: Test Settingmentioning

confidence: 99%

See 3 more Smart Citations

Skin cancer classification via convolutional neural networks: systematic review of studies involving human experts

Haggenmüller

Maron

Hekler

et al. 2021

European Journal of Cancer

Self Cite

181

113

View full text Add to dashboard Cite

show abstract

Section: Automated Skin Cancer Classification Of Dermoscopic Imagesmentioning

confidence: 99%

Section: Automated Skin Cancer Classification Of Dermoscopic Imagesmentioning

confidence: 99%

Section: Test Set Characteristicsmentioning

confidence: 99%

Section: Test Settingmentioning

confidence: 99%

See 2 more Smart Citations

Skin cancer classification via convolutional neural networks: systematic review of studies involving human experts

Haggenmüller

Maron

Hekler

et al. 2021

European Journal of Cancer

Self Cite

181

113

View full text Add to dashboard Cite

show abstract

“…Finally, dermatoscopic images are increasingly used for training of machine learning algorithms [70][71][72][73][74][75]. Computer algorithms based on deep learning outperformed dermatologists in some studies and increased the expectations that AI will replace human expertise, at least for some applications such as teledermatoscopy.…”

Section: Dermatoscopymentioning

confidence: 99%

Evolution of the Clinical, Dermoscopic and Pathologic Diagnosis of Melanoma

Kittler

2021

Dermatol Pract Concept

Self Cite

View full text Add to dashboard Cite

The conventional narrative states that the steadily rising incidence of melanoma among fair-skinned Caucasian populations during the last decades is caused by excessive UV-exposure. There is, however, no doubt that other factors had a significant impact on the rising incidence of melanoma. Pre-1980s the clinical diagnosis of melanoma was based on gross criteria such as ulceration or bleeding. Melanomas were often diagnosed in advanced stages when the prognosis was grim. In the mid-1980s education campaigns such as the propagation of the ABCD criteria, which addressed health care professionals and the public alike, shifted the focus towards early recognition. Dermatoscopy, which became increasingly popular in the mid-1990s, improved the accuracy for the diagnosis of melanoma in comparison to inspection with the unaided eye, especially for flat and small lesions lacking ABCD criteria. At the same time, pathologists began to lower their thresholds, particularly for the diagnosis of melanoma in situ. The melanoma epidemic that followed was mainly driven by an increase in the number of in situ or microinvasive melanomas. In a few decades, the landscape shifted from an undercalling to an overcalling of melanomas, a development that is now met with increased criticism. The gold standard of melanoma diagnosis is still conventional pathology, which is faced with low to moderate interobserver agreement. New insights in the molecular landscape of melanoma did not translate into techniques for the reliable diagnosis of gray zone lesions including small lesions. The aim of this review is to put our current view of melanoma diagnosis in historical context and to provide a narrative synthesis of its evolution. Based on this narrative I will provide suggestions on how to rebuild the trust in melanoma diagnosis accuracy and in the benefit of early recognition.

show abstract

Bilderkennung mittels künstlicher Intelligenz in der Hautkrebsdiagnostik

Winkler

Haenssle

2022

Dermatologie

View full text Add to dashboard Cite

Skin lesions of face and scalp – Classification by a market-approved convolutional neural network in comparison with 64 dermatologists

Cited by 31 publications

References 19 publications

Skin cancer classification via convolutional neural networks: systematic review of studies involving human experts

Skin cancer classification via convolutional neural networks: systematic review of studies involving human experts

Evolution of the Clinical, Dermoscopic and Pathologic Diagnosis of Melanoma

Bilderkennung mittels künstlicher Intelligenz in der Hautkrebsdiagnostik

Contact Info

Product

Resources

About