Pathologist-level classification of histopathological melanoma images with deep neural networks

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 25e26% of discordance for classifying a benign nevus versus malignant melanoma. Deep learning was successfully implemented to enhance the precision of lung and breast cancer diagnoses. The aim of this study is to illustrate the potential of deep learning to assist human assessment for a histopathologic melanoma diagnosis. Methods: Six hundred ninety-five lesions were classified by an expert histopathologist in accordance with current guidelines (350 nevi and 345 melanomas). Only the haematoxylin and eosin stained (H&E) slides of these lesions were digitalised using a slide scanner and then randomly cropped. Five hundred ninety-five of the resulting images were used for the training of a convolutional neural network (CNN). The additional 100 H&E image sections were used to test the results of the CNN in comparison with the original class labels.

Machine Learning in Dermatology: Current Applications, Opportunities, and Limitations

“…Géraud et al [74] found that 15% of the images used in the trial had no recognizable melanocytic lesion whatsoever, which indicates that these images only contained perilesional normal skin tissue. Hekler et al [77] also released a similar study a few months earlier with the same methods and model, claiming that their models achieved pathologist-level classification of histopathological melanoma images, but this study still has the same pitfalls.…”

Deep Learning in der Pathologie: Anwendungsmöglichkeiten und Herausforderungen in der Ophthalmopathologie

“…Eine ebenfalls häufige Fragestellung ist die Unterscheidung zwischen benignen und malignen melanozytären Tumoren. Hekler et al trainierten ein Deep-Learning-System auf die Unterscheidung zwischen benignen Nävi und malignen Melanomen [11]. Der Algorithmus zeigte in 19 % der Fälle eine Abweichung zum Standarduntersucher im Vergleich zu bis zu 25 % Abweichungen zwischen 2 verschiedenen Untersuchern [11].…”

Skin Lesions Classification Using Deep Learning Based on Dilated Convolution

“…To classify skin lessons, She et al [28] use several features like diameter, color, border, and asymmetry. To attain skin lesion segmentation, several pre-processing methods such as artificial removal, color transformation, lesion localization, contrast enhancement can be employed [29].…”