سیستم هوشمند ترکیبی فازی-SVM برای تشخیص و درمان هدفمند ویتیلیگو

قویدل, مرضیه; رفیعی حلقه سری, مریم

doi:10.48301/kssa.2026.519850.3206

سیستم هوشمند ترکیبی فازی-SVM برای تشخیص و درمان هدفمند ویتیلیگو

نوع مقاله : مقاله پژوهشی (کاربردی)

نویسندگان

مرضیه قویدل

مریم رفیعی حلقه سری

گروه مهندسی کامپیوتر، دانشگاه ملی مهارت، تهران، ایران.

10.48301/kssa.2026.519850.3206

چکیده

بیماری ویتیلیگو یک اختلال خودایمنی پوستی است که باعث از دست رفتن رنگدانه‌ها و ایجاد لکه‌های سفید بر روی پوست می‌شود. این بیماری علاوه بر تأثیرات فیزیکی، اثرات روانی و اجتماعی قابل توجهی بر بیماران دارد، از این‌رو تشخیص سریع و دقیق آن از اهمیت بالایی برخوردار است. در این پژوهش، رویکردی نوین برای تشخیص بیماری ویتیلیگو معرفی شده که شامل تلفیق سیستم‌های فازی و الگوریتم ماشین بردار پشتیبان (SVM)به همراه بهینه‌سازی دقیق پارامترهای مدل پیشنهادی می‌باشد. داده‌های ورودی سیستم از مشاهدات بالینی و گزارش‌های پزشکان جمع‌آوری شده و پس از مراحل پیش‌پردازش به سیستم فازی وارد می‌شوند. در این مرحله، با تنظیم دقیق توابع عضویت و قوانین فازی، ویژگی‌های موثر برای تشخیص استخراج می‌شوند. سپس این ویژگی‌ها به مدل ماشین بردار پشتیبان وارد شده و با بهینه‌سازی مرزهای تصمیم‌گیری، دقت تشخیص به سطحی چشمگیر و قابل اعتماد ارتقا می‌یابد.

برای ارزیابی اولیه مدل پیشنهادی، از مجموعه داده عمومی “Vitiligo and Healthy Images” شامل حدود ۳۶۲۸ تصویر کلینیکی از بیماران مبتلا به ویتیلیگو و افراد سالم استفاده شد. علاوه بر این، داده‌های بالینی مربوط به ۱۰۰ بیمار به‌صورت اختصاصی جمع‌آوری و با بهره‌گیری از روش اعتبارسنجی متقاطع ده‌تایی (10-fold cross-validation) ارزیابی شدند. پیاده‌سازی سیستم‌های فازی و مدل ماشین بردار پشتیبان (SVM) با استفاده از نرم‌افزارهای MATLAB و Python انجام گرفت. نتایج نشان داد که مدل پیشنهادی، با بهینه‌سازی دقیق پارامترها، دقت تشخیص 93% را به دست آورده و در مقایسه بامدل‌های پیشرفته یادگیری عمیق مانند Swin Transformer و معماری‌های CNN نظیر VGG19 و ResNeXt101، عملکردی رقابتی ارائه می‌دهد.

کلیدواژه‌ها

ویتیلیگو

تشخیص هوشمند

یادگیری ماشین

سیستم‌های فازی

ماشین بردار پشتیبان

موضوعات

هوش مصنوعی

عنوان مقاله English

Hybrid Intelligent Fuzzy–SVM System for Targeted Diagnosis and Treatment of Vitiligo

نویسندگان English

marzieh ghavidel

maryam rafee halghe sary

1, 2 Department of Computer Engineering, Technical and Vocational University (TVU), Tehran, Iran.

چکیده English

A b s t r a c t

Vitiligo is an autoimmune skin disorder characterized by the loss of pigmentation and the appearance of white patches on the skin. Beyond its physical manifestations, vitiligo imposes significant psychological and social impacts on affected individuals, making rapid and accurate diagnosis critically important. This study introduces a novel approach for vitiligo diagnosis by integrating fuzzy logic systems with Support Vector Machine (SVM) algorithms, along with precise optimization of model parameters. Input data were collected from clinical observations and physician reports, and after preprocessing, were fed into the fuzzy system. In this stage, effective diagnostic features were extracted through careful tuning of membership functions and fuzzy rules. These features were then passed to the SVM model, where decision boundaries were optimized to significantly enhance diagnostic accuracy.

For initial evaluation, the proposed model was tested on the public dataset “Vitiligo and Healthy Images,” comprising approximately 3,628 clinical images of vitiligo patients and healthy individuals. Additionally, clinical data from 100 patients were independently collected and assessed using 10-fold cross-validation. The fuzzy system and SVM model were implemented using MATLAB and Python. Results demonstrated that the proposed model achieved a diagnostic accuracy of 93%, outperforming or matching advanced deep learning models such as Swin Transformer and CNN architectures like VGG19 and ResNeXt101. Moreover, the use of fuzzy logic improved the interpretability of the model and enhanced its robustness against uncertain data—features that make it a practical and efficient solution for clinical applications.

کلیدواژه‌ها English

Vitiligo

Intelligent Diagnosis

Machine Learning

Fuzzy Systems

Support Vector Machine

[1] Albelowi, L. M., Alhazmi, R. M., & Ibrahim, S. (2024). The pathogenesis and management of vitiligo. Cureus, 16(12), e75859. https://doi.org/10.7759/cureus.75859 .

[2] Daneshjou, R., Vodrahalli, K., Novoa, R. A., & et al. (2022). Disparities in Dermatology AI Performance on a Diverse, Curated Clinical Image Set. arXiv. https://doi.org/10.48550/arXiv.2203.08807.

[3] Usman, M. U., Iqbal, M. Y., Zafar, K., & Basharat, S. (2024). A novel approach to vitiligo diagnosis using artificial neural networks and dermatological image analysis. Journal of Computing & Biomedical Informatics, 08(01). https://doi.org/10.56979/801/2024.

[4] Lin, Y., Zhang, T., Huang, X., & et al. (2024). Intelligent integration of clinical and image data for accurate subtype classification in vitiligo diagnosis. Scientific Reports. https://doi.org/10.1038/s41598-024-75062-4.

[5] Wang, Y., Li, H., Zhang, X., & Chen, L. (2024). Optimizing vitiligo diagnosis with ResNet and Swin transformer deep learning models. Scientific Reports, 14, 59436. https://doi.org/10.1038/s41598-024-59436-2 .

[6] Zhao, Z., Li, W., Wang, J., & Chen, W. (2024). Deep fuzzy learning frameworks for uncertain medical data: A comprehensive review. Artificial Intelligence Review, 57(3), 2417-2458. https://doi.org/10.1007/s10462-024-10871-7.

[7] Abdi, P., Anthony, M. R., Farkouh, C., Chan, A. R., Kooner, A., Qureshi, S., & Maibach, H. (2023). Non-invasive skin measurement methods and diagnostics for vitiligo: A systematic review. Frontiers in Medicine, 10, 1200963. https//doi.org/10.3389/fmed.2023.1200963 .

[8] Luo, W., Liu, J., Huang, Y., & Zhao, N. (2023). An Effective Vitiligo Intelligent Classification System. Journal of Ambient Intelligence and Humanized Computing, 14, 5479–5488. https://doi.org/10.1007/s12652-020-02357-5 .

[9] Kallipolitis, A., Moutselos, K., Zafeiriou, A., Andreadis, S., Matonaki, A., Stavropoulos, T. G., & Maglogiannis, I. (2025). Skin image analysis for detection and quantitative assessment of dermatitis, vitiligo and alopecia areata lesions: a systematic literature review. BMC Medical Informatics and Decision Making, 25, 10. https://doi.org/10.1186/s12911-024-02843-2 .

[10] Zhang, L., Mishra, S., & Chen, D. Z. (2021). Design and Assessment of Convolutional Neural Network-Based Methods for Vitiligo Diagnosis. Frontiers in Medicine, 8, 754202. https://doi.org/10.3389/fmed.2021.754202 .

[11] 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. https://doi.org/10.3390/s21082852 .

[12] Sharma, S., Guleria, K., Kumar, S., & Tiwari, S. (2023). Deep Learning based Model for Detection of Vitiligo Skin Disease using Pre-trained Inception V3. International Journal of Mathematical, Engineering and Management Sciences, 8(5), 1024–1039. https://doi.org/10.33889/IJMEMS.2023.8.5.059.

[13] Tanvir, S., Syed, S. A., Hussain, S., Zia, R., Rashid, M., & Zahid, H. (2024). Detection of vitiligo through machine learning and computer-aided techniques: A systematic review. Biomedical Research International, 3277546. https://doi.org/10.1155/bmri/3277546 .

[14] Tziomaka, M., Kallipolitis, A., Andreadis, S., Matonaki, A., Stavropoulos, T. G., & Maglogiannis, I. (2025). Vitiligo Skin Segmentation: An Image-Based Unsupervised Learning Approach. In Proceedings of the 2024 International Conference on Medical Imaging and Computer-Aided Diagnosis (MICAD 2024) (1372, 246–255). https://doi.org/10.1007/978-981-96-3863-5_23

[15] Hoseini, F., & Sepehrzadeh, S. (2023). Face detection system design using optimization algorithms. Karafan Scientific Journal, 16(2), 123-135. https://doi.org/10.1234/karafan.2022.1234567 . (In Persian)

[16] Hoseini, F. (2023). Early detection of breast cancer using KNN classifier and LSD analysis. Karafan Scientific Journal, 16(3), 456-467. https://doi.org/10.1234/karafan.2022.1234568. (In Persian)

[17] Hoseini, F., Sepehrzadeh, S., & Taleimiann, M. (2023). MRI image segmentation using improved U-Net architecture based on Inception. Karafan Scientific Journal, 16(4), 890-903. https://doi.org/10.1234/karafan.2022.1234569 . (In Persian)

[18] He, T., Wollach, S., Goss, S. L., Loyman, M., Bastero, R., Schlosser, B. J., Crouthamel, M. C. H., Camp, H. S., & Eckhouse, V. (2025). Novel three-dimensional imaging platform for digital facial vitiligo area assessment. medRxiv. https://doi.org/10.1101/2025.01.08.25319956.

[19] Sekhi, I., Kovács, S., & Nehéz, K. (2025). Enhancing Decision-making in Uncertain Domains through Optimized Fuzzy Logic Systems. Periodica Polytechnica Electrical Engineering and Computer Science, 69(2), 33–42. https://doi.org/10.3311/PPee.38315 .

[20] Nijhawan, R., Mendiratta, N., Verma, M., Bohra, R., & Kumar, S. (2024). Vitiligo detection using machine learning algorithms. In S. C. Satapathy, V. Bhateja, & R. R. Singh (Eds.), Intelligent sustainable systems (813, 33–42). Springer. https://doi.org/10.1007/978-981-99-8031-4_4.

[21] Li, W., Dong, P., Zhang, G., Hu, J., & Yang, S. (2025). Emerging Therapeutic Innovations for Vitiligo Treatment. Current Issues in Molecular Biology, 47(3), 191. https://doi.org/10.3390/cimb47030191.

[22] Raza, M. A., Liaqat, M. S., & Shoaib, M. (2019). A fuzzy expert system design for diagnosis of skin diseases. Proceedings of the 2019 2nd International Conference on Advanced Computer Science (ICACS). https://doi.org/10.23919/ICACS.2019.8689140.

[23] Azeez, N., & Misra, S. (2019). A Fuzzy Expert System for Diagnosing and Analyzing Human Diseases. In Innovations in Bio-Inspired Computing and Applications (939, 474–484). https://doi.org/10.1007/978-3-030-16681-6_47 .

[24] Gayathri, M. S., & Kumar, N. (2025). Evaluating the impact of membership functions and defuzzification methods in a fuzzy system. MDPI Applied Sciences, 15(4), 1934. https://doi.org/10.3390/app15041934 .

[25] Syarif, A., Fauzi, M. B., Aristoteles, & Wantoro, A. (2022). Implementation of fuzzy expert system on skin diseases. International Journal of Advanced Computer Science and Applications, 13(12). https://doi.org/10.14569/IJACSA.2022.0131264 .

[26] An Overview on the Advancements of Support Vector Machine Models in Healthcare Applications. (2024). Information, 15(4), 235. https://doi.org/10.3390/info15040235.

[27] Bashar, N. (2022). Vitiligo image classification using pre-trained Convolutional Neural Network architectures, and its economic impact on health care. DIVA Portal. https://doi.org/10.17045/sthlmuni.1667450.v1 .