Abstract:In total, 520 mucocutaneous melanocytic tumours, evenly divided between melanomas and naevi, were labelled only at the whole‐slide or whole‐tissue section level. Data were split 80% for training, 10% for validation and 10% for testing and three multiple‐instance learning frameworks were evaluated to predict melanoma or not. Stratified fivefold Monte Carlo cross‐validation was performed. The average AUC result for Trans‐MIL was 0.99, CLAM 0.99 and DTFD‐MIL 0.97. Background The broad histomorphological spectrum of melanocytic pathologies requires large data sets to develop accurate and generalisable deep learning (DL)‐based diagnostic pathology classifiers. Weakly supervised DL promotes utilisation of larger training data sets compared to fully supervised (patch annotation) approaches. Objectives To evaluate weakly supervised DL image classifiers for discriminating melanomas from naevi on haematoxylin and eosin (H&E)‐stained pathology slides. Methods A representative H&E slide for 260 naevi and 260 melanomas from mucocutaneous sites at one tertiary institution was digitized. Clinicopathological features were recorded for each case including thickness and histological subtype. Whole‐slide or whole‐tissue section labels were applied. The ground truth was established by consensus diagnosis from two pathologists. Multiple‐instance learning models, Trans‐MIL, CLAM and DTFD‐MIL were evaluated at 10×, 20× and 40× magnifications using stratified fivefold Monte Carlo cross‐validation, with 80/10/10 splits for training/validation/test groups, to predict melanoma from naevus. Heatmaps were generated to understand model performance. Results Naevi cases were younger (median age: 51 years; melanoma median age: 71.5 years), with more balanced sex distribution (males: 48.8%, melanoma male subgroup: 64.2%). The most frequent histological subtypes of naevi and melanomas were dysplastic compound (n = 99, 38.1%) and superficial spreading (n = 124, 47.7%), respectively. Average AUC (±1 SD) for Trans‐MIL, CLAM and DTFD‐MIL across test groups were 0.9952 ± 0.006, 0.9925 ± 0.0052 and 0.9708 ± 0.0328, at 20× magnification, respectively. Performance of the models varied according to the magnification used. Heatmaps from the two best performing models, Trans‐MIL and CLAM, generally indicated attention on appropriate tissue regions for interpretation. Conclusions Weakly supervised DL on pathological slides of common mucocutaneous melanocytic tumours provides highly accurate diagnostic value for discrimination of melanomas and naevi. External validation and further assessment on less frequently occurring histologic subtypes and borderline cases using this method is required.

Diagnostic assessment of deep learning for melanocytic lesions using whole-slide pathological images

Automated identification of malignancy in whole-slide pathological images: identification of eyelid malignant melanoma in gigapixel pathological slides using deep learning

[Deep learning-based fully automated intelligent and precise diagnosis for melanocytic lesions]

Deep Learning-Based Accurate Diagnosis of Eyelid Malignant Melanoma from Gigapixel Pathologic Slides

Deep Learning-Based Semantic Segmentation of Non-Melanocytic Skin Tumors in Whole-Slide Histopathological Images.

Region of Interest Detection in Melanocytic Skin Tumor Whole Slide Images-Nevus and Melanoma

Region of Interest Detection in Melanocytic Skin Tumor Whole Slide Images -- Nevus & Melanoma

Utilizing deep learning model for assessing melanocytic density in resection margins of lentigo maligna

Diagnostic Performance of Deep Learning Algorithms Applied to Three Common Diagnoses in Dermatopathology

Interpretable Diagnosis for Whole-Slide Melanoma Histology Images Using Convolutional Neural Network

Automated differentiation of skin melanocytes from keratinocytes in high‐resolution histopathology images using a weakly‐supervised deep‐learning framework

Deep learning model shows pathologist-level detection of sentinel node metastasis of melanoma and intra-nodal nevi on whole slide images

Detection and Localization of Melanoma Skin Cancer in Histopathological Whole Slide Images

Weakly supervised deep learning image analysis can differentiate melanoma from naevi on haematoxylin and eosin‐stained histopathology slides

Interpretable Classification from Skin Cancer Histology Slides Using Deep Learning: A Retrospective Multicenter Study

Evaluating deep learning-based melanoma classification using immunohistochemistry and routine histology: A three center study

Development and validation of a deep learning model for improving detection of nonmelanoma skin cancers treated with Mohs micrographic surgery

Deep Learning Techniques for the Dermoscopic Differential Diagnosis of Benign/Malignant Melanocytic Skin Lesions: From the Past to the Present

Unveiling the power of convolutional neural networks in melanoma diagnosis

A deep learning model based on whole slide images to predict disease-free survival in cutaneous melanoma patients