Literature DB >> 29846502

Man against machine: diagnostic performance of a deep learning convolutional neural network for dermoscopic melanoma recognition in comparison to 58 dermatologists.

H A Haenssle1, C Fink1, R Schneiderbauer1, F Toberer1, T Buhl2, A Blum3, A Kalloo4, A Ben Hadj Hassen5, L Thomas6, A Enk1, L Uhlmann7.   

Abstract

Background: Deep learning convolutional neural networks (CNN) may facilitate melanoma detection, but data comparing a CNN's diagnostic performance to larger groups of dermatologists are lacking.
Methods: Google's Inception v4 CNN architecture was trained and validated using dermoscopic images and corresponding diagnoses. In a comparative cross-sectional reader study a 100-image test-set was used (level-I: dermoscopy only; level-II: dermoscopy plus clinical information and images). Main outcome measures were sensitivity, specificity and area under the curve (AUC) of receiver operating characteristics (ROC) for diagnostic classification (dichotomous) of lesions by the CNN versus an international group of 58 dermatologists during level-I or -II of the reader study. Secondary end points included the dermatologists' diagnostic performance in their management decisions and differences in the diagnostic performance of dermatologists during level-I and -II of the reader study. Additionally, the CNN's performance was compared with the top-five algorithms of the 2016 International Symposium on Biomedical Imaging (ISBI) challenge.
Results: In level-I dermatologists achieved a mean (±standard deviation) sensitivity and specificity for lesion classification of 86.6% (±9.3%) and 71.3% (±11.2%), respectively. More clinical information (level-II) improved the sensitivity to 88.9% (±9.6%, P = 0.19) and specificity to 75.7% (±11.7%, P < 0.05). The CNN ROC curve revealed a higher specificity of 82.5% when compared with dermatologists in level-I (71.3%, P < 0.01) and level-II (75.7%, P < 0.01) at their sensitivities of 86.6% and 88.9%, respectively. The CNN ROC AUC was greater than the mean ROC area of dermatologists (0.86 versus 0.79, P < 0.01). The CNN scored results close to the top three algorithms of the ISBI 2016 challenge. Conclusions: For the first time we compared a CNN's diagnostic performance with a large international group of 58 dermatologists, including 30 experts. Most dermatologists were outperformed by the CNN. Irrespective of any physicians' experience, they may benefit from assistance by a CNN's image classification. Clinical trial number: This study was registered at the German Clinical Trial Register (DRKS-Study-ID: DRKS00013570; https://www.drks.de/drks_web/).

Entities:  

Mesh:

Year:  2018        PMID: 29846502     DOI: 10.1093/annonc/mdy166

Source DB:  PubMed          Journal:  Ann Oncol        ISSN: 0923-7534            Impact factor:   32.976


  154 in total

1.  Reliable test of clinicians' mastery in skin cancer diagnostics.

Authors:  Niels Kvorning Ternov; T Vestergaard; L Rosenkrantz Hölmich; K Karmisholt; A L Wagenblast; H Klyver; M Hald; L Schøllhammer; L Konge; A H Chakera
Journal:  Arch Dermatol Res       Date:  2020-06-28       Impact factor: 3.017

2.  Design of an Algorithm for Automated, Computer-Guided PASI Measurements by Digital Image Analysis.

Authors:  Christine Fink; Tobias Fuchs; Alexander Enk; Holger A Haenssle
Journal:  J Med Syst       Date:  2018-11-03       Impact factor: 4.460

Review 3.  [Artificial intelligence in cardiology : Relevance, current applications, and future developments].

Authors:  Bettina Zippel-Schultz; Carsten Schultz; Dirk Müller-Wieland; Andrew B Remppis; Martin Stockburger; Christian Perings; Thomas M Helms
Journal:  Herzschrittmacherther Elektrophysiol       Date:  2021-01-15

Review 4.  Designing deep learning studies in cancer diagnostics.

Authors:  Andreas Kleppe; Ole-Johan Skrede; Sepp De Raedt; Knut Liestøl; David J Kerr; Håvard E Danielsen
Journal:  Nat Rev Cancer       Date:  2021-01-29       Impact factor: 60.716

5.  Artificial Intelligence and the Softer Side of Medicine.

Authors:  Joseph A Craft
Journal:  Mo Med       Date:  2018 Sep-Oct

6.  The empowerment of all modalities against cancer.

Authors:  Shuting Han; Han Chong Toh
Journal:  Singapore Med J       Date:  2018-11       Impact factor: 1.858

7.  Multiphoton microscopy of the dermoepidermal junction and automated identification of dysplastic tissues with deep learning.

Authors:  Mikko J Huttunen; Radu Hristu; Adrian Dumitru; Iustin Floroiu; Mariana Costache; Stefan G Stanciu
Journal:  Biomed Opt Express       Date:  2019-12-10       Impact factor: 3.732

8.  Expert-Level Diagnosis of Nonpigmented Skin Cancer by Combined Convolutional Neural Networks.

Authors:  Philipp Tschandl; Cliff Rosendahl; Bengu Nisa Akay; Giuseppe Argenziano; Andreas Blum; Ralph P Braun; Horacio Cabo; Jean-Yves Gourhant; Jürgen Kreusch; Aimilios Lallas; Jan Lapins; Ashfaq Marghoob; Scott Menzies; Nina Maria Neuber; John Paoli; Harold S Rabinovitz; Christoph Rinner; Alon Scope; H Peter Soyer; Christoph Sinz; Luc Thomas; Iris Zalaudek; Harald Kittler
Journal:  JAMA Dermatol       Date:  2019-01-01       Impact factor: 10.282

Review 9.  Melanoma Early Detection: Big Data, Bigger Picture.

Authors:  Tracy Petrie; Ravikant Samatham; Alexander M Witkowski; Andre Esteva; Sancy A Leachman
Journal:  J Invest Dermatol       Date:  2018-10-25       Impact factor: 8.551

10.  Hidden Stratification Causes Clinically Meaningful Failures in Machine Learning for Medical Imaging.

Authors:  Luke Oakden-Rayner; Jared Dunnmon; Gustavo Carneiro; Christopher Ré
Journal:  Proc ACM Conf Health Inference Learn (2020)       Date:  2020-04
View more

北京卡尤迪生物科技股份有限公司 © 2022-2023.