Accuracy of a Smartphone-Based Artificial Intelligence Application for Classification of Melanomas, Melanocytic Nevi, and Seborrheic Keratoses

Current artificial intelligence algorithms can classify melanomas at a level equivalent to that of experienced dermatologists. The objective of this study was to assess the accuracy of a smartphone-based “You Only Look Once” neural network model for the classification of melanomas, melanocytic nevi, and seborrheic keratoses. The algorithm was trained using 59,090 dermatoscopic images. Testing was performed on histologically confirmed lesions: 32 melanomas, 35 melanocytic nevi, and 33 seborrheic keratoses. The results of the algorithm’s decisions were compared with those of two skilled dermatologists and five beginners in dermatoscopy. The algorithm’s sensitivity and specificity for melanomas were 0.88 (0.71–0.96) and 0.87 (0.76–0.94), respectively. The algorithm surpassed the beginner dermatologists, who achieved a sensitivity of 0.83 (0.77–0.87). For melanocytic nevi, the algorithm outclassed each group of dermatologists, attaining a sensitivity of 0.77 (0.60–0.90). The algorithm’s sensitivity for seborrheic keratoses was 0.52 (0.34–0.69). The smartphone-based “You Only Look Once” neural network model achieved a high sensitivity and specificity in the classification of melanomas and melanocytic nevi with an accuracy similar to that of skilled dermatologists. However, a bigger dataset is required in order to increase the algorithm’s sensitivity for seborrheic keratoses.

[1]  H. Haenssle,et al.  Assessment of Diagnostic Performance of Dermatologists Cooperating With a Convolutional Neural Network in a Prospective Clinical Study: Human With Machine. , 2023, JAMA dermatology.

[2]  A. Lallas,et al.  European consensus-based interdisciplinary guideline for melanoma. Part 1: Diagnostics: Update 2022. , 2022, European journal of cancer.

[3]  P. Tschandl,et al.  Validation of artificial intelligence prediction models for skin cancer diagnosis using dermoscopy images: the 2019 International Skin Imaging Collaboration Grand Challenge , 2022, The Lancet. Digital health.

[4]  F. Bray,et al.  Global Burden of Cutaneous Melanoma in 2020 and Projections to 2040. , 2022, JAMA dermatology.

[5]  N. Hart,et al.  Dermoscopy use in primary care: a qualitative study with general practitioners , 2022, BMC Primary Care.

[6]  J. Delhommelle,et al.  Machine Learning and Deep Learning Algorithms for Skin Cancer Classification from Dermoscopic Images , 2022, Bioengineering.

[7]  T. Nijsten,et al.  Validation of a Market-Approved Artificial Intelligence Mobile Health App for Skin Cancer Screening: A Prospective Multicenter Diagnostic Accuracy Study , 2022, Dermatology.

[8]  Jakob Nikolas Kather,et al.  Skin cancer classification via convolutional neural networks: systematic review of studies involving human experts. , 2021, European journal of cancer.

[9]  M. Romanelli,et al.  Seborrheic keratosis-like melanoma: a diagnostic challenge , 2021, Melanoma research.

[10]  H. Haenssle,et al.  Collective human intelligence outperforms artificial intelligence in a skin lesion classification task , 2021, Journal der Deutschen Dermatologischen Gesellschaft = Journal of the German Society of Dermatology : JDDG.

[11]  H. Haenssle,et al.  Dark corner artefact and diagnostic performance of a market‐approved neural network for skin cancer classification , 2021, Journal der Deutschen Dermatologischen Gesellschaft = Journal of the German Society of Dermatology : JDDG.

[12]  K. Meiburger,et al.  The Role in Teledermoscopy of an Inexpensive and Easy-to-Use Smartphone Device for the Classification of Three Types of Skin Lesions Using Convolutional Neural Networks , 2021, Diagnostics.

[13]  Y. Yang,et al.  A deep learning-based hybrid artificial intelligence model for the detection and severity assessment of vitiligo lesions , 2021, Annals of translational medicine.

[14]  R. Hofmann-Wellenhof,et al.  Skin lesions of face and scalp - Classification by a market-approved convolutional neural network in comparison with 64 dermatologists. , 2020, European journal of cancer.

[15]  T. Maurer,et al.  Cost Minimization Analysis of a Teledermatology Triage System in a Managed Care Setting. , 2020, JAMA dermatology.

[16]  V. de Giorgi,et al.  Folliculotropism in head and neck lentigo maligna and lentigo maligna melanoma , 2020, Journal der Deutschen Dermatologischen Gesellschaft = Journal of the German Society of Dermatology : JDDG.

[17]  George Shih,et al.  A patient-centric dataset of images and metadata for identifying melanomas using clinical context , 2020, Scientific Data.

[18]  Sumit Kumar Singh,et al.  Melanoma Diagnosis Using Deep Learning and Fuzzy Logic , 2020, Diagnostics.

[19]  John Paoli,et al.  Human–computer collaboration for skin cancer recognition , 2020, Nature Medicine.

[20]  Benjamin Izar,et al.  Effects of Label Noise on Deep Learning-Based Skin Cancer Classification , 2020, Frontiers in Medicine.

[21]  T. Nijsten,et al.  Accuracy of a smartphone application for triage of skin lesions based on machine learning algorithms , 2019, Journal of the European Academy of Dermatology and Venereology : JEADV.

[22]  A. Kalloo,et al.  Computer Algorithms Show Potential for Improving Dermatologists' Accuracy to Diagnose Cutaneous Melanoma; Results of ISIC 2017. , 2020, Journal of the American Academy of Dermatology.

[23]  R. Hofmann-Wellenhof,et al.  Man against machine reloaded: performance of a market-approved convolutional neural network in classifying a broad spectrum of skin lesions in comparison with 96 dermatologists working under less artificial conditions. , 2020, Annals of oncology : official journal of the European Society for Medical Oncology.

[24]  T. Maurer,et al.  Impact of Teledermatology on the Accessibility and Efficiency of Dermatology Care in an Urban Safety-Net Hospital: A Pre-Post analysis. , 2019, Journal of the American Academy of Dermatology.

[25]  R. Hofmann-Wellenhof,et al.  Evaluation of dermoscopic criteria for seborrheic keratosis on non‐polarized versus polarized dermoscopy , 2019, Skin research and technology : official journal of International Society for Bioengineering and the Skin (ISBS) [and] International Society for Digital Imaging of Skin (ISDIS) [and] International Society for Skin Imaging.

[26]  S. Arias-Santiago,et al.  Proportion of Potentially Avoidable Referrals From Primary Care to Dermatologists for Cystic Lesions or Benign Neoplasms in Spain: Analysis of Data From the DIADERM Study , 2019, Actas Dermo-Sifiliográficas (English Edition).

[27]  Enes Ayan,et al.  Skin Lesion Segmentation in Dermoscopic Images with Combination of YOLO and GrabCut Algorithm , 2019, Diagnostics.

[28]  K. Liopyris,et al.  Dermoscopy: A Review of the Structures That Facilitate Melanoma Detection , 2019, The Journal of the American Osteopathic Association.

[29]  Achim Hekler,et al.  Deep learning outperformed 136 of 157 dermatologists in a head-to-head dermoscopic melanoma image classification task. , 2019, European journal of cancer.

[30]  Silvio Savarese,et al.  Generalized Intersection Over Union: A Metric and a Loss for Bounding Box Regression , 2019, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[31]  Noel C. F. Codella,et al.  Skin Lesion Analysis Toward Melanoma Detection 2018: A Challenge Hosted by the International Skin Imaging Collaboration (ISIC) , 2019, ArXiv.

[32]  H. Haenssle,et al.  Man against machine: diagnostic performance of a deep learning convolutional neural network for dermoscopic melanoma recognition in comparison to 58 dermatologists , 2018, Annals of oncology : official journal of the European Society for Medical Oncology.

[33]  A. Sober,et al.  Contrasting features of childhood and adolescent melanomas , 2018, Pediatric dermatology.

[34]  Harald Kittler,et al.  The HAM10000 dataset, a large collection of multi-source dermatoscopic images of common pigmented skin lesions , 2018, Scientific Data.

[35]  A. Kalloo,et al.  Results of the 2016 International Skin Imaging Collaboration International Symposium on Biomedical Imaging challenge: Comparison of the accuracy of computer algorithms to dermatologists for the diagnosis of melanoma from dermoscopic images , 2018, Journal of the American Academy of Dermatology.

[36]  M. I. Fernandez,et al.  Examining the factors associated with past and present dermoscopy use among family physicians , 2017, Dermatology practical & conceptual.

[37]  Sebastian Thrun,et al.  Dermatologist-level classification of skin cancer with deep neural networks , 2017, Nature.

[38]  Kilian Q. Weinberger,et al.  Densely Connected Convolutional Networks , 2016, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[39]  Noel C. F. Codella,et al.  Skin lesion analysis toward melanoma detection: A challenge at the 2017 International symposium on biomedical imaging (ISBI), hosted by the international skin imaging collaboration (ISIC) , 2016, 2018 IEEE 15th International Symposium on Biomedical Imaging (ISBI 2018).

[40]  Ali Farhadi,et al.  You Only Look Once: Unified, Real-Time Object Detection , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[41]  S. Menzies,et al.  Increase in the sensitivity for melanoma diagnosis by primary care physicians using skin surface microscopy , 2000, The British journal of dermatology.

[42]  J. R. Landis,et al.  The measurement of observer agreement for categorical data. , 1977, Biometrics.

[43]  S. Arias-Santiago,et al.  Proportion of Potentially Avoidable Referrals From Primary Care to Dermatologists for Cystic Lesions or Benign Neoplasms in Spain: Analysis of Data from the DIADERM Study. , 2019, Actas dermo-sifiliograficas.