OSHQOZON TASVIRINI BIPOLYAR NORAVSHAN QAYTA ISHLASH ORQALI TASHXISLASHNING GIBRID ALGORITMI
Keywords:
Bipolyar noravshan mantiq, tibbiy tasvirni qayta ishlash, oshqozon endoskopiyasi, BF-HybNet arxitekturasi, gibrid neyron tarmoqlar, A_VAR muammolari, poliplarni aniqlash, DuS-KFCM klasterlash, GLCM tekstura xususiyatlari, Self-Attention mexanizmi, Diagnostik noaniqlik, BUWL (Bipolyar Uncertainty Weighted Loss)Abstract
Ushbu maqolada oshqozon tasvirlarini avtomatlashtirilgan tarzda tahlil qilish uchun yangi yondashuv – bipolyar noravshan mantiqqa asoslangan dastlabki ishlov berish va gibrid neyron tarmoq arxitekturasini birlashtirgan Sinergetik model taklif etiladi. Mazkur metodologiya tasvirlarni noaniqlikdan xabardor xususiyatlar maydoniga aylantirib, chuqur o‘rganish uchun kerakli belgilarni samarali ajratib olishga yordam beradi. Bipolyar noravshan mantiqqa asoslangan dastlabki ishlov berish bosqichi tasvirlardagi sun’iy artefaktlarni yumshatadi va klassik gastroskopiyada ko‘rinmaydigan o‘ta nozik diagnostika belgilarini ajratib ko‘rsatadi. Masalan, Gray-Level Co-Occurrence Matrix (GLCM) orqali tekstura xususiyatlarini aniqlash va noravshan klasterlash usullari oshqozondagi qon ketish holatlarini segmentatsiya qilishda yuqori ko‘rsatgich ko‘rsatdi. Taklif etilgan neyron tarmog‘i va uning termoyadroviy moduli kichik poliplarni murakkab fonlardan aniqlashda an’anaviy yondashuvlarga nisbatan yuqori aniqlik beradi. Ayniqsa, saraton xavfi 10–20% atrofida bo‘lgan adenomatoz poliplarni aniqlashda ushbu yondashuv samarali natijalar beradi.Model maxsus yig‘ilgan 1433 ta o‘quv va 508 ta test gastroskopik tasvirlar to‘plamida sinovdan o‘tkazildi. Sinov natijalariga ko‘ra, aniqlik 91,6%, eslab qolish darajasi 86,2%, F1 balli 88,8% va F2 balli 87,2% ni tashkil etdi. Umumiy aniqlik 88,7% bo‘ldi.Mazkur yondashuv oshqozon tasvirlarini avtomatlashtirilgan va aniq baholash imkonini berib, diagnostik noaniqliklarni kamaytiradi va klinik belgilarni ertaroq va ishonchli tarzda aniqlashda amalga oshirishga xizmat qiladi.
References
1. Al-Husban, A., Amourah, A., & Jaber, J. J. (2020). Bipolar complex fuzzy sets and their properties. *Italian Journal of Pure and Applied Mathematics, 43*, 756–767. https://ijpam.uniud.it/online_issue/202043/63%20Husbana-Amourah-Jaber.pdf
2. Selvaraj, N., & Sundararajan, R. (2012). An efficient filtering technique for denoising medical images. *Procedia Engineering, 38*, 3680–3686. https://doi.org/10.1016/j.proeng.2012.06.424
3. Yang, Y., Liu, X., Wang, L., & Nie, F. (2021). Fuzzy logic in medical image processing: A review. *IEEE Reviews in Biomedical Engineering, 14*, 314–328. https://doi.org/10.1109/RBME.2020.3014084
4. Jahan, T., & Ahmad, T. (2016). Adaptive fuzzy logic approach for image segmentation. *Procedia Computer Science, 85*, 666–673. https://doi.org/10.1016/j.procs.2016.05.270
5. Selvaraj, N., & Raja, K. B. (2020). A survey on fuzzy clustering techniques in medical image segmentation. *Materials Today: Proceedings, 33*, 2676–2682. https://doi.org/10.1016/j.matpr.2020.04.672
6. Xu, Z., & Yager, R. R. (2006). Intuitionistic and bipolar fuzzy aggregation functions. *Fuzzy Sets and Systems, 157(9)*, 1140–1154. https://doi.org/10.1016/j.fss.2005.11.011
7. Ma, J., & Lu, Y. (2023). Vision transformers and their applications in medical imaging: A survey. *Computerized Medical Imaging and Graphics, 102*, 102185. https://doi.org/10.1016/j.compmedimag.2022.102185
8. Liu, Z., Mao, H., Wu, C. Y., Feichtenhofer, C., Darrell, T., & Xie, S. (2022). A ConvNet for the 2020s. In *Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)*, 11976–11986. https://doi.org/10.1109/CVPR52688.2022.01169
9. Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., ... & Polosukhin, I. (2017). Attention is all you need. In *Advances in Neural Information Processing Systems (NeurIPS), 30*. https://papers.nips.cc/paper_files/paper/2017/file/3f5ee243547dee91fbd053c1c4a845aa-Paper.pdf
10. Li, X., Chen, H., Qi, X., Dou, Q., Fu, C. W., & Heng, P. A. (2018). H-DenseUNet: Hybrid densely connected UNet for liver and tumor segmentation from CT volumes. *IEEE Transactions on Medical Imaging, 37(12)*, 2663–2674. https://doi.org/10.1109/TMI.2018.2845918
11. Chen, T., & Guestrin, C. (2016). XGBoost: A scalable tree boosting system. In *Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining* (pp. 785–794). https://doi.org/10.1145/2939672.2939785
12. He, K., Zhang, X., Ren, S., & Sun, J. (2016). Deep residual learning for image recognition. In *Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR)*, 770–778. https://doi.org/10.1109/CVPR.2016.90
Downloads
Published
Issue
Section
License
Copyright (c) 2025 Iskandarova Sayyora Nurmamatovna, Abdurashidova Kamola Turg’unbayevna, Kuchkarov Temurbek Ataxanovich, Sabitova Nazokat Qobuljon qizi
This work is licensed under a Creative Commons Attribution 4.0 International License.
