Skin wounds are a common medical concern, and their early detection and accurate classification are essential for timely intervention and effective treatment. Manual assessment is often time-consuming, subjective, and prone to errors, highlighting the need for automated, reliable solutions. In this study, we propose a deep learning-based framework for automated binary and multiclass classification of skin wounds from clinical images. The model integrates MobileNetV2 for feature extraction, applies binary classification to distinguish normal versus wounded skin, and performs multiclass classification to identify specific wound types. Additionally, a Grad-CAM-based interpretability module provides visual explanations, and the system is deployed via a Streamlit application for real-time clinical testing. Experimental results demonstrate high accuracy and robust performance, validating the model’s potential for rapid and reliable wound assessment. Future enhancements include incorporating larger datasets, multimodal inputs, and predictive analytics for improved generalization and clinical utility.

This study successfully developed a MobileNetV2-based wound detection system achieving 96.82% binary and 90.81% multiclass validation accuracy across nine wound types. The integration of Grad-CAM provides interpretable visual explanations, while Streamlit deployment enables real-time clinical use. Future enhancements include expanding dataset diversity across skin tones, implementing longitudinal healing analysis, integrating multimodal sensor data, and conducting multi-center clinical validation studies to establish regulatory approval pathways.

[1] Carrión, H., Jafari, M., Bagood, M. D., Yang, H. Y., Isseroff, R. R., & Gomez, M. (2022). Automatic wound detection and size estimation using deep learning algorithms. PLoS Computational Biology, 18(3), e1009852. [2] Wang, C., Anisuzzaman, D. M., Williamson, V., Dhar, M. K., Rostami, B., Niezgoda, J., Gopalakrishnan, S., & Yu, Z. (2020). Fully automatic wound segmentation with deep convolutional neural networks. Scientific Reports, 10(1), 21897. [3] Jahangir, M. Z. B., Akter, S., Nasim, M. D., Gupta, K. D., & George, R. (2024). Deep learning for automated wound classification and segmentation. arXiv preprint arXiv:2408.11064. [4] Zhang, R., Tian, D., Xu, D., Qian, W., & Yao, Y. (2022). A survey of wound image analysis using deep learning: classification, detection, and segmentation. IEEE Access, 10, 79502-79515. [5] Buschi, D., Curti, N., Cola, V., Carlini, G., Sala, C., Dall’Olio, D., … & Giampieri, E. (2023). Automated wound image segmentation: transfer learning from human to pet via active semi-supervised learning. Animals, 13(6), 956. [6] Muhtasim, D. A., Pavel, M. I., Paul, A., Shakib, A. W., Gazi, S. S., & Hasan, M. M. (2024). Automated wound segmentation using attention mechanism based on enhanced MobileNetV2. 2024 IEEE 3rd International Conference on Robotics, Automation, Artificial-Intelligence and Internet-of-Things (RAAICON), 96-101. [7] Srinivasu, P. N., SivaSai, J. G., Ijaz, M. F., Bhoi, A. K., Kim, W., & Kang, J. J. (2021). Classification of skin disease using deep learning neural networks with MobileNet V2 and LSTM. Sensors, 21(8), 2852. [8] Maulana, S. A., & Syahputra, H. (2025). Image classification of wounds in diabetes mellitus patients using convolutional neural network (CNN) algorithm. Journal of Artificial Intelligence and Engineering Applications (JAIEA), 5(1), 1909-1916. [9] Monroy, B., Bacca, J., Sanchez, K., Arguello, H., & Castillo, S. (2021). Two-step deep learning framework for chronic wounds detection and segmentation: A case study in Colombia. 2021 XXIII Symposium on Image, Signal Processing and Artificial Vision (STSIVA), 1-6. [10] Jeribi, F., Siddiqa, A., Kibriya, H., Tahir, A., & Rana, N. (2025). Efficient wound classification using YOLO11n: A lightweight deep learning approach. Computers, Materials & Continua, 85(1). [11] Firasari, E., Khasanah, N., Cahyanti, F. L. D., Kholifah, D. N., Khultsum, U., & Sarasati, F. (2024). Performance evaluation of ResNet50 and MobileNetV2 in skin cancer image classification with various optimizers. 2024 International Conference on Information Technology Research and Innovation (ICITRI), 376-380. [12] Eldem, H., Ülker, E., & Işıklı, O. Y. (2022). Classification of pressure and diabetic chronic wound tissue images with deep learning methods. [13] Narayanan, H., & Ghanta, S. (2024). A study of data augmentation techniques to overcome data scarcity in wound classification using deep learning. arXiv preprint arXiv:2411.02456. [14] Jahangir, M. Z. B., Akter, S., Nasim, M. D., Gupta, K. D., & George, R. (2024). Deep learning for automated wound classification and segmentation. arXiv preprint arXiv:2408.11064. [15] Anisuzzaman, D. M., Patel, Y., Niezgoda, J., Gopalakrishnan, S., & Yu, Z. (2022). Wound severity classification using deep neural network. arXiv preprint arXiv:2204.07942. [16] Lin, Y., Lien, Y., Lin, C., & Shen, S. (2025). The application of a deep learning image recognition model based on multi-scale convolutional attention mechanism for clinical open wound detection. 2025 1st International Conference on Consumer Technology (ICCT-Pacific), 1-4. [17] Scebba, G., Zhang, J., Catanzaro, S., Mihai, C., Distler, O., Berli, M., & Karlen, W. (2021). Detect-and-Segment: a deep learning approach to automate wound image segmentation. Informatics in Medicine Unlocked, 100884. [18] Ramachandram, D., Ramírez-GarcíaLuna, J., Fraser, R., Martínez-Jiménez, M., Arriaga-Caballero, J., & Allport, J. (2022). Fully automated wound tissue segmentation using deep learning on mobile devices: Cohort study. JMIR mHealth and uHealth, 10, e36977. [19] Haval, A., Mishra, A., & Dash, S. (2024). Smart wound monitoring and healing assessment system with deep learning methods – A systematic review. Journal of Angiotherapy, 8(1), 1-12. [20] Patel, Y., Shah, T., Dhar, M., Zhang, T., Niezgoda, J., Gopalakrishnan, S., & Yu, Z. (2023). Integrated image and location analysis for wound classification: a deep learning approach. Scientific Reports, 14, 4567. [21] Faria, J., Almeida, J., Vasconcelos, M., & Rosado, L. (2019). Automated mobile image acquisition of skin wounds using real-time deep neural networks. International Conference on Image Analysis and Processing, 61-73. [22] Rajkumar, M., Siriyala, S., Sharma, A., & Kumar, D. (2024). Comparative analysis of ResNet, EfficientNet, and MobileNetV2 for detecting skin conditions in deep learning models. 2024 International Conference on Sustainable Communication Networks and Application (ICSCNA), 1487-1492. [23] Zheng, X., Yang, Z., Sutarlie, L., Thangaveloo, M., Yu, Y., Salleh, N., Chin, J., Xiong, Z., Becker, D., Loh, X., Tee, B., & Su, X. (2023). Battery-free and AI-enabled multiplexed sensor patches for wound monitoring. Science Advances, 9, eadg6670. [24] Huang, S., Chu, Y., Liu, L., Yao, W., Chen, Y., Yu, C., Yu, C., Tung, K., Chiu, H., & Tsai, M. (2023). Deep learning-based clinical wound image analysis using a Mask R-CNN architecture. Journal of Medical and Biological Engineering, 43, 417-426. [25] Muhammad Adnan et al. (2023). Journal of Physics: Conference Series, 2547, 012005. [26] Cassidy, B., Kendrick, C., Reeves, N. D. et al. (2025). Deep learning in chronic wound segmentation: a comprehensive review and meta-analysis. The Visual Computer, 41, 11885-11908.

© 2025 International Journal of Engineering and Techniques (IJET).

Digital Object Identifier (DOI)

IJET assigns a unique DOI to every accepted article via Zenodo for persistent linking and citation. Your article’s DOI: https://doi.org/{{doi}}

Open DOI About Zenodo DOI

Close