Oʻzbekcha
MIYA O‘SIMTALARINI MASHINALI O‘QITISH VA CHUQUR O‘QITISH USULLARI YORDAMIDA TIBBIY TASVIRLARDAN YASHIRIN PATOLOGIK BELGILARNI ANIQLASH
Journal
Priority areas for applying artificial intelligence to pedagogical education
Issue
Priority areas for applying artificial intelligence to pedagogical education
Abstract
Mazkur maqolada CNN, U-Net, Vision Transformer va transfer learning usullari yordamida rentgen, KT, MRT tasvirlaridan patologiyalarni aniqlash tadqiq etilgan. Gibrid CNN-Transformer modeli NIH ChestX-Ray14 to‘plamida AUC = 0.972, BraTS 2023 da Dice = 0.91 va LUNA16 da CPM = 0.891 ko‘rsatkichlariga erishdi. FastAPI va TensorFlow/Keras asosida miya MRI tasvirlari uchun real vaqtli diagnostika tizimi ishlab chiqilib, 96.8% aniqlikka erishildi. Grad-CAM++ mexanizmi klinik ishonchlilikni ta’minlaydi
Keywords
chuqur o‘qitish
miya o‘simtasi
konvolyutsion neyron tarmoq
MRI
FastAPI
TensorFlow
Grad-CAM++
Glioma
Meningioma
Pituitary tumor.
Русский
В данной работе исследуется обнаружение патологий на рентгеновских снимках, КТ и МРТ с использованием методов CNN, U-Net, Vision Transformer и трансферного обучения. Гибридная модель CNN-Transformer достигла AUC = 0,972 на наборе данных NIH ChestX-Ray14, Dice = 0,91 на BraTS 2023 и CPM = 0,891 на LUNA16. Была разработана система диагностики в реальном времени для изображений МРТ головного мозга на основе FastAPI и TensorFlow/Keras, достигшая точности 96,8%. Движок Grad-CAM++ обеспечивает клиническую надежность.
глубокое обучение
опухоль головного мозга
сверточная нейронная сеть
МРТ
FastAPI
TensorFlow
Grad-CAM++
глиома
менингиома
опухоль гипофиза.
English
This paper investigates the detection of pathologies from X-ray, CT, and MRI images using CNN, U-Net, Vision Transformer, and transfer learning methods. The hybrid CNN-Transformer model achieved AUC = 0.972 on the NIH ChestX-Ray14 dataset, Dice = 0.91 on BraTS 2023, and CPM = 0.891 on LUNA16. A real-time diagnostic system for brain MRI images based on FastAPI and TensorFlow/Keras was developed, achieving 96.8% accuracy. The Grad-CAM++ engine ensures clinical reliability.
deep learning
brain tumor
convolutional neural network
MRI
FastAPI
TensorFlow
Grad-CAM++
Glioma
Meningioma
Pituitary tumor
1. Esteva, A. et al. (2019). A guide to deep learning in healthcare. Nature Medicine, 25, 24–29.
2. Litjens, G. et al. (2017). A survey on deep learning in medical image analysis. Medical Image Analysis, 42, 60–88.
3. Litjens, G. et al. (2017). Ibid.
4. Ronneberger, O., Fischer, P., & Brox, T. (2015). U-Net: CNNs for Biomedical Image Segmentation. MICCAI.
5. Havaei, M. et al. (2017). Brain tumor segmentation with deep neural networks. Medical Image Analysis, 35, 18–31.
6. Dosovitskiy, A. et al. (2021). An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale. ICLR.
7. Selvaraju, R.R. et al. (2020). Grad-CAM: Visual Explanations from Deep Networks. IJCV, 128, 336–359.
8. Tajbakhsh, N. et al. (2016). CNNs for medical image analysis: Full training or fine tuning? IEEE TMI, 35(5), 1299–1312.
9. CBTRUS Statistical Report: Primary Brain and Other CNS Tumors. (2023). Neuro-Oncology, 25(Suppl 4).
10. Kaggle Brain MRI Images Dataset. [Online]. Available: https://www.kaggle.com/datasets/navoneel/brain-mri-images-for-brain-tumor-detection
11. D.K. Muxamediyeva, A.A. Abdiraximov (2025) “Miya oʻsimtalarini erta tashxislashda sun’iy intellektdan foydalanish: backend api asosidagi yondashuv” “Образование наука и инновационные идеи в мире» Xalqaro ilmiy elektron jurnal” 3-tom 69-soni ISSN 2181-3187:176-182 https://scientific-jl.com/obr/issue/view/227
12. D.K. Muxamediyeva, A.A. Abdiraximov (2025) “Miya oʻsimtalarini mri va kt tasvirlar toʻplamlarini shakllantirish hamda oldindan ishlov berish” “TECHSCIENCE.UZ - Texnika fanlarining dolzarb masalalari” ilmiy elektron jurnal” 3-jild 3-soni ISSN:3030-3702 6-12 https://techscience.uz/index.php/jurnal/issue/view/8
13. M.Y. Doshanova, A.A. Abdiraximov (2025) “Data augmentation va transfer learning strategiyalarining miya o‘simtalarini tashxislashdagi qo‘llanilishi va sun’iy intellekt modelarining samaradorligiga ta’siri” “Development of science – Fan taraqqiyoti” ilmiy elektron jurnal” 1-jild 8-soni ISSN:3030-3907 154-159 https://devos.uz/article.php?id=1517