828 Dermatopathologist-level classification of skin cancer with deep neural networks at multi-magnification

Abstract

Conventional monotherapies only benefit a minority of melanoma patients while combined immunotherapy exhibited extremely high rates of treatment-related adverse events. Herein, we created a multifunctional and immunogenic nanoplatform, AuNR@mSiO 2 @DOX-Cu x S-PEG, which integrating photothermal properties of gold nanoparticles, photodynamic properties of CuxS and chemotherapy into a single nanoplatform. Upon near-infrared laser irradiation (NIR), the Cu x S were uncapped and triggered chemotherapy drugs release into tumor environment. The properties and cytotoxicity to melanoma cells were explored in vitro. In addition, the anti-tumor effect and molecular mechanism of inducing tumor-specific immune responses were also explored in vitro and in vivo. The nanoplatform was stabilized, facilely synthesized and exhibited a high drug payload up to 36.37AE2.74%. Compared to control groups, the nanoplatform could forcefully inhibit tumor growth in situ (p<0.001) and lung metastasis with lower time and power intensity. Furthermore, it could promote maturation of dendritic cells and induce the infiltration of both CD8 + and CD4 + T lymphocytes in tumor in vivo. Mechanistically, the nanoplatform not only can eradicate cancer cells by inducing cell apoptosis and/or necrosis directly, but also induce a strong tumor-specific immune response by resulting in calreticulin exposure on the surface of cancer cell (p<0.01) and released highmobility group box 1 (HMGB1) protein (p<0.01). Our study developed a promising NIRtriggered multifunctional and immunogenic nanocomposite, which could induce anti-tumor immunity and would be a potential strategy for melanoma in advanced stage. 828Dermatopathologist-level classification of skin cancer with deep neural networks at multi-magnification Artificial intelligence technologies, as novel diagnostic tools, are now affecting medical

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

“…In these methods, various features (preferably those which are important for a particular scenario) are extracted through either deep learning or neural networks by feeding large datasets along with the corresponding classification labels [8,9]. Diagnostic convolutional neural networks (CNN) have matched or exceeded the expected ability of field experts in several pathological image recognition tasks [10,11] particularly for the diagnosis of the lung and breast cancer at the earliest possible state [12,13]. Likewise, in the skin pathology recognition task, Hekler et al [14] have demonstrated the pathologist-level classification of malignant melanomas versus benign nevi using a pretrained ResNet50 CNN.…”

Image analysis of cutaneous melanoma histology: a systematic review and meta-analysis

“…Overall, 4,888 specimens were included, of which at least 2,715 were melanoma specimens. The diagnostic entities within the datasets varied between studies, with some only containing melanoma deposits 12,21,[24][25][26]33 and others containing more than one pathology 10,11,22,[27][28][29][30][31][32] .…”

“…There was between-study variation in terms of intended use of the IA. Most studies focused on a binary classification task, with some focussing on detection and localisation of melanoma deposits in WSIs containing melanoma (melanoma versus not melanoma) 12,25,26,32 and others performing diagnostic classifications including melanomas versus naevi 11,22,31 and primary melanoma versus metastatic melanoma 33 . Five studies addressed more complex classifications into three or more diagnostics entities 10,23,[28][29][30] .…”

“…There was some degree of variation in the IT used. Most employed the use of a convolutional neural networks (CNN), with the architecture differing considerably [10][11][12]22,23,26,[29][30][31]33 . Studies using CNNs were more recently conducted.…”

“…There was heterogeneity in the reference standard. In most studies pathologists provided diagnostic labels [10][11][12]22,31 , categorised specimens by histological features 23 , carried out manual annotation 21,25,28,30 or interpreted immunohistochemical staining 23,24,26 . Some studies used a combination of these approaches 23,32 .…”