Skin cancer high-risk patient screening from dermoscopic images via Artificial Intelligence: an online study

Abstract

Objectives: To evaluate a novel Artificial Intelligence (AI) method for the detection of malignant skin lesions from dermoscopic images. Methods: 58,457 dermoscopic images available online from the International Skin Imaging Collaboration (ISIC) Archive were downloaded. These images were acquired from different centers worldwide by recognized dermatologists and show varied clinical outcomes belonging to different types of benign and malign skin lesions. A state-of-the-art AI skin lesion classifier based on Deep Learning was designed. The method, fully automated, first locates and segments the nevus in the image and then classifies it into either benign or malign type. Results: 1,631 images (2.8%) were discarded due to bad quality. A total of 56,826 images were finally used. Two thirds of the images (37,688) were used to train the classifier, leaving the remaining 19,138 images for validation. In this set, malignant lesions had a prevalence of 15.4% (2,956/19,138). The AI skin lesion classifier reached an area under the curve (AUC) of 87.4%. Optimal cut-off point in terms of accuracy resulted in an 85.9% accuracy (16,439/19,138) and sensibility of 89.6% (2,648/ 2,956) at 85.2% (13,791/16,182) specificity. Negative predictive value (NPV) was 97.8% (13,791/14,099). Other training/validation splits were also evaluated, showing similar results. Conclusions: A novel AI method showed promising results as skin lesion classifier from dermoscopic images. Its high NPV value could make it suited for high-risk patient screening. A large clinical study to confirm these results is needed and will be pursued.

Efficacy of a Deep Learning Convolutional Neural Network System for Melanoma Diagnosis in a Hospital Population

“…After the retraining, the quantusSKIN obtained better values of sensitivity (from 0.655 to 0.691), accuracy (from 0.767 to 0.776), and the rest of the parameters, except specificity. This fact would show that CNN models should not only be trained with images from public databases [ 11 ], but also from the medical consultations where they are used, considering, for example, the different skin phenotypes in each medical population. In fact, the influence of skin color on the efficacy of the quantusSKIN system was not investigated in the current study, which supposed one of its main limitations.…”

“…Originally, the quantusSKIN system was trained with 37,688 training images obtained from the public database ISIC Archive 2019 and 2020 [ 11 ]. In this current study, this network was retrained with the additional 339 training images (196 nevi and 143 melanomas) of 291 different patients obtained from the clinical practice of the Hospital Universitario Ramón y Cajal.…”

“…As reported in a previous study, our research group trained a deep learning CNN with 37,688 biopsy-verified images from the International Skin Imaging Collaboration (ISIC) Archive to classify images into either benign or malignant skin lesions, to reach acceptable sensitivity and specificity values [ 11 ]. Nevertheless, no previous study focused on the use of a deep learning-based method as screening tools in primary health care rather than diagnosis techniques in specialized dermatology consultation.…”

An Improved Skin Lesion Classification Using a Hybrid Approach with Active Contour Snake Model and Lightweight Attention-Guided Capsule Networks

“…A novel and trustworthy feature fusion model for skin cancer identification was proposed [ 64 ]. First, the images are cleaned of noise using a Gaussian filter (GF).…”