Abstract

Automated methods for dermoscopy image analysis support the de¬tection of skin cancer in an early stage and reducing skin cancer-related deaths. Deep learning based methods have shown com¬pelling results for the analysis of skin cancer images. In this paper, an enhanced ECRNet-based hybrid model is explored for the clas¬sification and detection of skin cancer and diverse skin anomalies. This model utilizes an ensemble of classification models, namely, ResNet50, ResNet101, MobileNetV2, Vision Transformer, Con-vNeXt, DeiT-Small, EL-DLOA, WavIntNet, Conformer, Xception, VGG16 and an Ensemble of ECRNet and other models. For le¬sion localization, the YOLO model family and Faster R-CNN ar¬chitecture are examined. Experimental results demonstrate that the Hybrid-Ensemble approach outperforms other models with an ac¬curacy of 97.2%, precision of 95.4%, recall of 93.7%, and F1 score of 94.5% while YOLOV26 achieved an mAP of 71.3% with a pre¬cision of 73.9%. Grad-CAM was used to improve the model inter¬pretability by highlighting the image regions containing the lesions. A web application for image upload, automated prediction, and diagnostic visualization was developed using Flask and SQLite.

References

• Halil Murat U¨ nver and Enes Ayan. Skin lesion segmentation in dermoscopic images with combination of yolo and grabcut algorithm. Diagnostics, 9(3):72, 2019.
• DD Yang, JD Salciccioli, DC Marshall, A Sheri, and J Shal-houb. Trends in malignant melanoma mortality in 31 coun¬tries from 1985 to 2015. British Journal of Dermatology, 183(6):1056–1064, 2020.
• Li Hu, SF Jin, Ling Chen, and YL Wang. Trends in the in¬cidence and mortality of cutaneous melanoma in hong kong between 1983 and 2015. Int J Clin Exp Med, 11(8):8259–8266, 2018.
• Alfonso Baldi, Marco Quartulli, Raffaele Murace, Emanuele Dragonetti, Mario Manganaro, Oscar Guerra, and Stefano Bizzi. Automated dermoscopy image analysis of pigmented skin lesions. Cancers, 2(2):262–273, 2010.
• Yuexiang Li and Linlin Shen. Skin lesion analysis towards melanoma detection using deep learning network. Sensors, 18(2):556, 2018.
• RM Lucas, S Yazar, AR Young, M Norval, FR De Gruijl, Y Takizawa, LE Rhodes, CA Sinclair, and RE Neale. Human health in relation to exposure to solar ultraviolet radiation un¬der changing stratospheric ozone and climate. Photochemical & Photobiological Sciences, 18(3):641–680, 2019.
• Arthur AM Teodoro, Douglas H Silva, Renata L Rosa, Muhammad Saadi, Lunchakorn Wuttisittikulkij, Rao Asad Mumtaz, and Demostenes Z Rodriguez. A skin cancer clas¬sification approach using gan and roi-based attention mecha¬nism. Journal of Signal Processing Systems, 95(2):211–224, 2023.
• Chao Xin, Zhifang Liu, Keyu Zhao, Linlin Miao, Yizhao Ma, Xiaoxia Zhu, Qiongyan Zhou, Songting Wang, Lingzhi Li, Feng Yang, et al. An improved transformer network for skin cancer classification. Computers in Biology and Medicine, 149:105939, 2022.
• Soumyya Kanti Datta, Mohammad Abuzar Shaikh, Sargur N Srihari, and Mingchen Gao. Soft attention improves skin can¬cer classification performance. In International Workshop on Interpretability of Machine Intelligence in Medical Image Computing, pages 13–23. Springer, 2021.
• Md Shahin Ali, Md Sipon Miah, Jahurul Haque, Md Mah-bubur Rahman, and Md Khairul Islam. An enhanced tech¬nique of skin cancer classification using deep convolutional neural network with transfer learning models. Machine Learning with Applications, 5:100036, 2021.
• Luzhou Liu, Xiaoxia Zhang, and Zhinan Xu. An adaptive weight search method based on the grey wolf optimizer al-gorithm for skin lesion ensemble classification. International Journal of Imaging Systems and Technology, 34(2):e23049, 2024.
• Xin Shi, Xiaofang Dong, Shuhua Ye, Wei Li, and Haiyan Li. Wavelet integrated multiscale feature fusion network for im¬balanced skin lesion classification. 2024.
• Sobia Bibi, Muhammad Attique Khan, Jamal Hussain Shah, Robertas Damasˇevicˇius, Areej Alasiry, Mehrez Marzougui, Majed Alhaisoni, and Anum Masood. Msrnet: Multiclass skin lesion recognition using additional residual block based fine-tuned deep models information fusion and best feature selec¬tion. Diagnostics, 13(19):3063, 2023.
• Veena Dillshad, Muhammad Attique Khan, Muhammad Nazir, Oumaima Saidani, Nazik Alturki, and Seifedine Kadry. D2lfs2net: Multi-class skin lesion diagnosis using deep learn¬ing and variance-controlled marine predator optimisation: An application for precision medicine. CAAI Transactions on In¬telligence Technology, 10(1):207–222, 2025.
• Muneezah Hussain, Muhammad Attique Khan, Robertas Damasˇevicˇius, Areej Alasiry, Mehrez Marzougui, Majed Al¬haisoni, and Anum Masood. Skinnet-inio: multiclass skin le¬sion localization and classification using fusion-assisted deep neural networks and improved nature-inspired optimization algorithm. Diagnostics, 13(18):2869, 2023.
• Ra Karthik, Tejas Sunil Vaichole, Sanika Kiran Kulkarni, Ojaswa Yadav, and Faiz Khan. Eff2net: An efficient chan¬nel attention-based convolutional neural network for skin dis¬ease classification. Biomedical Signal Processing and Con¬trol, 73:103406, 2022.
• Mehak Arshad, Muhammad Attique Khan, Usman Tariq, Ammar Armghan, Fayadh Alenezi, Muhammad Younus Javed, Shabnam Mohamed Aslam, and Seifedine Kadry. A computer-aided diagnosis system using deep learn¬ing for multiclass skin lesion classification. Computational intelligence and neuroscience, 2021(1):9619079, 2021.
• T Fikrle and K Pizinger. Digital computer analysis of dermatoscopical images of 260 melanocytic skin lesions; perimeter/area ratio for the differentiation between malig-nant melanomas and melanocytic nevi. Journal of the Eu-ropean Academy of Dermatology and Venereology, 21(1):48–55, 2007.
• Noel CF Codella, David Gutman, M Emre Celebi, Brian Helba, Michael A Marchetti, Stephen W Dusza, Aadi Kalloo, Konstantinos Liopyris, Nabin Mishra, Harald Kittler, et al. Skin lesion analysis toward melanoma detection: A challenge at the 2017 international symposium on biomedical imag¬ing (isbi), hosted by the international skin imaging collabo¬ration (isic). In 2018 IEEE 15th international symposium on biomedical imaging (ISBI 2018), pages 168–172. IEEE, 2018.
• Lucia Ballerini, Robert B Fisher, Ben Aldridge, and Jonathan Rees. A color and texture based hierarchical k-nn approach to the classification of non-melanoma skin lesions. In Color medical image analysis, pages 63–86. Springer, Berlin, Hei¬delberg, 2013.
• Yue Cheng, Ragavendar Swamisai, Scott E Umbaugh, Randy H Moss, William V Stoecker, Saritha Teegala, and Subhashini K Srinivasan. Skin lesion classification using rel¬ative color features. Skin Research and Technology, 14(1):53–64, 2008.
• Philipp Tschandl, Cliff Rosendahl, and Harald Kittler. The ham10000 dataset, a large collection of multi-source dermato-scopic images of common pigmented skin lesions. Scientific data, 5(1):180161, 2018.