RECOGNITION OF VIRAL DISEASES AND SKIN NEOPLASMS USING CONVOLUTIONAL NEURAL NETWORKS
Keywords:
convolutional neural network, classification, object detection, image classification, machine training, decision-making system, intelligent technology, optimisation, segmentationAbstract
This study focuses on the use of convolutional neural networks for the automated recognition of skin diseases, specifically targeting two main classes: skin neoplasms and viral diseases. Each of these classes contains a large number of subclasses with different etiologies and clinical manifestations, making their accurate differentiation challenging even for specialists. The goal of this research is not only to classify the main disease groups but also to analyze their subclasses to improve the accuracy of diagnostic algorithms.The neural network was trained using DermNet, one of the largest open dermatology image datasets, which contains a wide range of skin pathologies. The initial dataset was divided into training and test sets and supplemented with additional images from publicly available sources.This approach allowed the model to account for the variability in skin types, as well as differences in tone, texture, and lesion characteristics, which are critical factors in diagnosis.The model is based on deep learning, utilizing classical convolutional layers and additional normalization methods. It is capable of analyzing and classifying diseases from images with high accuracy. Final results showed an accuracy of 91%, a recall of 89%, and an F1-score of 90%, significantly surpassing previous results, which ranged between 0.7 and 0.75. This demonstrates that the application of advanced algorithms and adaptive training significantly enhances the capabilities of automated diagnostics.Unlike previous approaches that relied on basic convolutional models with minimal hyperparameter adjustments, this study has developed a more flexible system that not only classifies images by primary disease categories but also considers differences between disease subtypes. This is particularly important for distinguishing similar pathologies, such as benign and malignant neoplasms or different types of viral infections.A promising direction for future research is the development of a hierarchical neural network system, which will allow the model to first determine the general disease class and then further refine its classification into subtypes within each category. This approach will not only increase diagnostic accuracy but also provide a recommendation system for medical professionals regarding further patient examination. Additionally, there are plans to expand the dataset to test the model on other types of skin diseases, including rare pathological forms.
References
R. Sadik, A. Majumder, A. A. Biswas, B. Ahammad, and M. M. Rahman, “An in-depth analysis of Convolutional Neural Network architectures with transfer learning for skin disease diagnosis,” Healthcare Analytics, vol. 3, 2023, doi: 10.1016/j. health.2023.100143.
S. Karthikeyan, A. Kingsly Jabakumar, B. Anuradha, “Studies on Different CNN Algorithms for Face Skin Disease Classification Based on Clinical Image,” Proceeding International Conference on Science and Engineering, vol. 11, no. 1, 2023, doi: 10.52783/cienceng.v11i1.88.
F. D. Wibowo, I. Palupi, and B. A. Wahyudi, “Image Detection for Common Human Skin Diseases in Indonesia Using CNN and Ensemble Learning Method,” Journal of Computer System and Informatics (JoSYC), vol. 3, no. 4, 2022, doi: 10.47065/josyc. v3i4.2151.
S. Ahmed et al., “Human Skin Diseases Detection and Classification using CNN,” in 3rd International Conference on Electrical, Computer and Communication Engineering, ECCE 2023, 2023. doi: 10.1109/ECCE57851.2023.10101636.
A. Esteva et al., “Dermatologist-level classification of skin cancer with deep neural networks,” Nature, vol. 542, no. 7639, 2017, doi: 10.1038/nature21056.
B. Durgabhavani, B. Chandana, Sandhyarani, G. Lavanya, G. Srikanth, and N. Bhaskar, “Classification of Skin Disease using CNN,” in 2023 International Conference on Evolutionary Algorithms and Soft Computing Techniques, EASCT 2023, 2023. doi: 10.1109/EASCT59475.2023.10393477.
E. Kinshakov and Y. Parfenenko, “Application of Machine Learning Techniques to Solve the Problem of Skin Diseases Diagnosis,” in Studies in Systems, Decision and Control, vol. 496, Springer Science and Business Media Deutschland GmbH, 2023, pp. 101–116. doi: 10.1007/978-3-031-40997-4_7.
A. M. A. Talab, Z. Huang, F. Xi, and L. Haiming, “Detection crack in image using Otsu method and multiple filtering in image processing techniques,” Optik (Stuttg), vol. 127, no. 3, 2016, doi: 10.1016/j.ijleo.2015.09.147.
M. Kazemimoghadam, Z. Yang, M. Chen, L. Ma, W. Lu, and X. Gu, “Leveraging global binary masks for structure segmentation in medical images,” Phys Med Biol, vol. 68, no. 18, 2023, doi: 10.1088/1361-6560/acf2e2.
V. A. Ashwath, O. K. Sikha, and R. Benitez, “TS-CNN: A Three-Tier Self-Interpretable CNN for Multi-Region Medical Image Classification,” IEEE Access, vol. 11, 2023, doi: 10.1109/ACCESS.2023.3299850.
M. Frid-Adar, I. Diamant, E. Klang, M. Amitai, J. Goldberger, and H. Greenspan, “GAN-based synthetic medical image augmentation for increased CNN performance in liver lesion classification,” Neurocomputing, vol. 321, 2018, doi: 10.1016/j. neucom.2018.09.013.
K. Nirmala, K. Saruladha, and K. Dekeba, “Investigations of CNN for Medical Image Analysis for Illness Prediction,” Comput Intell Neurosci, vol. 2022, 2022, doi: 10.1155/2022/7968200.
L. Nanni, A. Manfè, G. Maguolo, A. Lumini, and S. Brahnam, “High performing ensemble of convolutional neural networks for insect pest image detection,” Ecol Inform, vol. 67, 2022, doi: 10.1016/j.ecoinf.2021.101515.
U. Kuran and E. C. Kuran, “Parameter selection for CLAHE using multiobjective cuckoo search algorithm for image contrast enhancement,” Intelligent Systems with Applications, vol. 12, 2021, doi: 10.1016/j.iswa.2021.200051.
Y. Patel et al., “An Improved Dense CNN Architecture for Deepfake Image Detection,” IEEE Access, vol. 11, 2023, doi: 10.1109/ACCESS.2023.3251417.
G. Cheng et al., “Prototype-CNN for Few-Shot Object Detection in Remote Sensing Images,” IEEE Transactions on Geoscience and Remote Sensing, vol. 60, 2022, doi: 10.1109/TGRS.2021.3078507.
