Abstract
Skin cancer is a major public health concern and the most common type of cancer among the other types. Reliable automated classification systems will provide clinicians with great help to detect malignant skin lesions as quickly as possible. Recently, deep learning-based approaches have efficiently outperformed other conventional machine learning models in medical image classification tasks. In this study, a novel computer-aided approach is designed for Skin Lesion Detection by creating an Ensemble of Deep (SLDED) models. More specifically, we initially performed a modified faster R-CNN using VGGNet feature extractor on ISIC archive database, including 4668 skin lesion images for lesion localization, and we obtained a mean average precision (mAP) of 0.96. Then we fused four different convolutional neural networks (CNNs) into one framework to obtain high classification accuracy. Moreover, a weighted majority voting method is proposed to aggregate the final decision of each individual voter. We evaluate our experimental classification results on 934 and 200 images from ISIC and PH2 test data. We achieved the average accuracy of 97.1% and 96%, Area under receiver operating characteristics curve (AUC) of 98.6% and 98.1%, precision of 87.1% and 90.2%, recall of 86.7% and 85.4% for ISIC and PH2 test data, respectively. As another objective evaluation, we have tested our proposed procedure on official test set of 2016 and 2017 International Symposium on Biomedical Imaging (ISIB) challenges. It outperforms the results of other proposed frameworks that have been published in those challenges. The results demonstrate that our proposed SLDED method is a meaningful approach to classify four different skin lesions with a high accuracy despite the lack of access to expensive computational equipment.
Similar content being viewed by others
Data availability
The datasets analyzed during the current study are available in the International Skin Imaging Collaboration (ISIC) Archive repository, https://challenge.isic-archive.com/data/.
References
Abbes W, Sellami D (2017) Automatic skin lesions classification using ontology-based semantic analysis of optical standard images. Procedia Comput Sci 112:2096–2105
Agarwal M, Damaraju N, Chaieb S Skin lesion analysis toward melanoma detection
Argenziano G, Soyer HP (2001) Dermoscopy of pigmented skin lesions–a valuable tool for early. Lancet Oncol 2(7):443–449
Argenziano G et al (2006) Dermoscopy improves accuracy of primary care physicians to triage lesions suggestive of skin cancer. J Clin Oncol 24(12):1877–1882
Attia M et al (2017) Skin melanoma segmentation using recurrent and convolutional neural networks. In: 2017 IEEE 14th International Symposium on Biomedical Imaging (ISBI 2017). IEEE
Badrinarayanan V, Kendall A, Cipolla R (2017) Segnet: A deep convolutional encoder-decoder architecture for image segmentation. IEEE Trans Pattern Anal Mach Intell 39(12):2481–2495
Barata C, Celebi ME, Marques JS (2017) Development of a clinically oriented system for melanoma diagnosis. Pattern Recogn 69:270–285
Baumann LS et al (2018) Safety and efficacy of hydrogen peroxide topical solution, 40%(w/w), in patients with seborrheic keratoses: results from 2 identical, randomized, double-blind, placebo-controlled, phase 3 studies (A-101-SEBK-301/302). J Am Acad Dermatol 79(5):869–877
Bengio Y (2009) Learning deep architectures for AI. Found Trends Mach Learn 2(1):1–127
Bi L et al (2017) Semi-automatic skin lesion segmentation via fully convolutional networks. In: 2017 IEEE 14th International Symposium on Biomedical Imaging (ISBI 2017). IEEE
Brinker TJ et al (2019) A convolutional neural network trained with dermoscopic images performed on par with 145 dermatologists in a clinical melanoma image classification task. Eur J Cancer 111:148–154
Brinker TJ et al (2019) Comparing artificial intelligence algorithms to 157 German dermatologists: the melanoma classification benchmark. Eur J Cancer 111:30–37
Brinker TJ et al (2019) Deep neural networks are superior to dermatologists in melanoma image classification. Eur J Cancer 119:11–17
Burdick J et al (2017) The impact of segmentation on the accuracy and sensitivity of a melanoma classifier based on skin lesion images. In: SIIM 2017 scientific program: Pittsburgh, PA, June 1-June 3, 2017, David L. Lawrence Convention Center
Carli P et al (2000) Preoperative assessment of melanoma thickness by ABCD score of dermatoscopy. J Am Acad Dermatol 43(3):459–466
Celebi ME et al (2007) A methodological approach to the classification of dermoscopy images. Comput Med Imaging Graph 31(6):362–373
center, c.s. Estimated new cases, 2019. Available from: https://cancerstatisticscenter.cancer.org/#!/
Chang W-Y et al (2013) Computer-aided diagnosis of skin lesions using conventional digital photography: a reliability and feasibility study. PLoS ONE 8(11):e76212
Codella NC et al (2017) 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). In: 2018 IEEE 15th International Symposium on Biomedical Imaging (ISBI 2018). IEEE
Collaboration, I.S.I. (2020) ISIC archive. Available from: https://www.isic-archive.com/#!/topWithHeader/wideContentTop/main
Conn AR, Gould NI, Toint P (1991) A globally convergent augmented Lagrangian algorithm for optimization with general constraints and simple bounds. SIAM J Numer Anal 28(2):545–572
Dascalu A, David E (2019) Skin cancer detection by deep learning and sound analysis algorithms: A prospective clinical study of an elementary dermoscope. EBioMedicine 43:107–113
Díaz IG (2017) Incorporating the knowledge of dermatologists to convolutional neural networks for the diagnosis of skin lesions. arXiv preprint arXiv:1703.01976
Esteva A et al (2017) Dermatologist-level classification of skin cancer with deep neural networks. Nature 542(7639):115
Goodfellow I, Bengio Y, Courville A (2016) Deep learning. MIT press
Grichnik JM, Rhodes AR, Sober AJ (2008) Benign neoplasias and hyperplasias of melanocytes. Fitzpatrick’s dermatology in general medicine, 7th edn, pp 1099–103
Haenssle HA et al (2018) Man against machine: diagnostic performance of a deep learning convolutional neural network for dermoscopic melanoma recognition in comparison to 58 dermatologists. Ann Oncol 29(8):1836–1842
Hahnloser RH et al (2000) Digital selection and analogue amplification coexist in a cortex-inspired silicon circuit. Nature 405(6789):947
Harangi B (2018) Skin lesion classification with ensembles of deep convolutional neural networks. J Biomed Inform 86:25–32
He K et al (2016) Deep residual learning for image recognition. In: Proceedings of the IEEE conference on computer vision and pattern recognition
He K et al (2017) Mask r-cnn. In: Proceedings of the IEEE international conference on computer vision
Hekler A et al (2019) Pathologist-level classification of histopathological melanoma images with deep neural networks. Eur J Cancer 115:79–83
Hekler A et al (2019) Superior skin cancer classification by the combination of human and artificial intelligence. Eur J Cancer 120:114–121
Hu H et al (2018) CNNAuth: continuous authentication via two-stream convolutional neural networks. In: 2018 IEEE international conference on networking, architecture and storage (NAS). IEEE
Isasi AG, Zapirain BG, Zorrilla AM (2011) Melanomas non-invasive diagnosis application based on the ABCD rule and pattern recognition image processing algorithms. Comput Biol Med 41(9):742–755
Jain S, Pise N (2015) Computer aided melanoma skin cancer detection using image processing. Procedia Comput Sci 48:735–740
Jaleel JA, Salim S, Aswin R (2013) Computer aided detection of skin cancer. In: 2013 International Conference on Circuits, Power and Computing Technologies (ICCPCT). IEEE
Kallenberg M et al (2016) Unsupervised deep learning applied to breast density segmentation and mammographic risk scoring. IEEE Trans Med Imaging 35(5):1322–1331
Kingma DP, Ba J (2014) Adam: a method for stochastic optimization. arXiv preprint arXiv:1412.6980
Krizhevsky A, Sutskever I, Hinton GE (2012) Imagenet classification with deep convolutional neural networks. In: Advances in neural information processing systems
Levine AB et al (2019) Rise of the machines: advances in deep learning for cancer diagnosis. Trends Cancer 5:157–169
Li Y, Hu H, Zhou G (2018) Using data augmentation in continuous authentication on smartphones. IEEE Internet Things J 6(1):628–640
Li Y et al (2020) Using feature fusion strategies in continuous authentication on smartphones. IEEE Internet Comput 24(2):49–56
Lopez AR et al (2017) Skin lesion classification from dermoscopic images using deep learning techniques. In: 2017 13th IASTED international conference on biomedical engineering (BioMed). IEEE
Majtner T, Yildirim-Yayilgan S, Hardeberg JY (2016) Combining deep learning and hand-crafted features for skin lesion classification. In: 2016 Sixth International Conference on Image Processing Theory, Tools and Applications (IPTA). IEEE
Manning C, Raghavan P, Schütze H (2010) Introduction to information retrieval. Nat Lang Eng 16(1):100–103
Mar VJ, Scolyer RA, Long GV (2017) Computer-assisted diagnosis for skin cancer: have we been outsmarted? Lancet 389(10083):1962–1964
Marchetti MA et al (2018) 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. J Am Acad Dermatol 78(2):270-277. e1
Maron RC et al (2019) Systematic outperformance of 112 dermatologists in multiclass skin cancer image classification by convolutional neural networks. Eur J Cancer 119:57–65
Matsunaga K et al (2017) Image classification of melanoma, nevus and seborrheic keratosis by deep neural network ensemble. arXiv preprint arXiv:1703.03108
Megahed M et al (2002) Reliability of diagnosis of melanoma in situ. Lancet 359(9321):1921–1922
Mendonca T et al (2015) PH2: A public database for the analysis of dermoscopic images. In: Dermoscopy image analysis. CRC Press
Menegola A et al (2017) RECOD titans at ISIC challenge 2017. arXiv preprint arXiv:1703.04819
Mirzaalian-Dastjerdi H et al (2018) Detecting and measuring surface area of skin lesions, in Bildverarbeitung für die Medizin 2018. Springer, pp 29–34
Moss RH et al (1989) Skin cancer recognition by computer vision. Comput Med Imaging Graph 13(1):31–36
Mueller SA et al (2019) Mutational patterns in metastatic cutaneous squamous cell carcinoma. J Invest Dermatol 139(7):1449-1458.e1
Nida N et al (2019) Melanoma lesion detection and segmentation using deep region based convolutional neural network and fuzzy C-means clustering. Int J Med Informatics 124:37–48
Okur E, Turkan M (2018) A survey on automated melanoma detection. Eng Appl Artif Intell 73:50–67
Renzi M et al (2019) Management of skin cancer in the elderly. Dermatol Clin 37(3):279–286
Sarıgül M, Avci BMOM (2019) Differential convolutional neural network. Neural Netw 116:279–287
Schaefer G et al (2014) An ensemble classification approach for melanoma diagnosis. Memetic Comput 6(4):233–240
Soudani A, Barhoumi W (2019) An image-based segmentation recommender using crowdsourcing and transfer learning for skin lesion extraction. Expert Syst Appl 118:400–410
Sreekantaswamy S et al (2019) Aging and the treatment of basal cell carcinoma. Clin Dermatol 37:373–378
Stoecker WV, Moss RH (1992) Digital imaging in dermatology. Elsevier
Stoecker WV et al (2005) Detection of asymmetric blotches (asymmetric structureless areas) in dermoscopy images of malignant melanoma using relative color. Skin Res Technol 11(3):179–184
Szegedy C et al (2015) Going deeper with convolutions. In: Proceedings of the IEEE conference on computer vision and pattern recognition
Uijlings JR et al (2013) Selective search for object recognition. Int J Comput Vision 104(2):154–171
Vasconcelos CN, Vasconcelos BN (2017) Experiments using deep learning for dermoscopy image analysis. Pattern Recognit Lett 139:95–103
Xie S et al (2017) Aggregated residual transformations for deep neural networks. In: Proceedings of the IEEE conference on computer vision and pattern recognition
Yu L et al (2016) Automated melanoma recognition in dermoscopy images via very deep residual networks. IEEE Trans Med Imaging 36(4):994–1004
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Competing interests
The authors have no competing interests in this study.
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Springer Nature or its licensor holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Shahsavari, A., Khatibi, T. & Ranjbari, S. Skin lesion detection using an ensemble of deep models: SLDED. Multimed Tools Appl 82, 10575–10594 (2023). https://doi.org/10.1007/s11042-022-13666-6
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11042-022-13666-6