SUN’IY INTELLEKT YORDAMIDA AKUSTIK SIGNALLARNI ANIQLIK BILAN O‘LCHASH VA KLASSIFIKATSIYA QILISH TIZIMI
DOI:
https://doi.org/10.65164/f29czx93Kalit so‘zlar:
sun’iy intellekt, akustik signal, chuqur o‘rganish, fonokardiografiya, mel- spektrogramma, klassifikatsiya, neyron tarmoqlar, signal qayta ishlash.Abstrak
Maqolada tibbiyot va sanoatda qo‘llaniladigan akustik signallarni qayta ishlash
uchun sun’iy intellekt asosida qurilgan avtomatlashtirilgan tizimning metodologiyasi taqdim
etilgan. Taklif etilgan yondashuv chuqur o‘rganish (deep learning) arxitekturasasida, xususan,
konvolyutsion neyron tarmoqlar (CNN) va uzun qisqa muddatli xotira (LSTM) tarmoqlarini
kombinatsiyalash orqali akustik signallarni real vaqtda tahlil qilish imkonini beradi. Tizim melspektrogramma
va MFCC (Mel-frequency cepstral coefficients) xususiyatlarini ajratib olish orqali
signalning zich va vaqtli xususiyatlarini aniqlaydi. Toshkent shahridagi tibbiyot muassasalaridan
to‘plangan 5,600 ta fonokardiografik va bronxoakustik yozuvlar asosida o‘tkazilgan sinovlar
natijasida tizimning aniqlik ko‘rsatkichi (accuracy) 94.2%, sezuvchanlik (sensitivity) 92.8% va
o‘zgarmaslik (specificity) 95.4% ga yetdi. Taklif etilgan yechim “Aqlli shifoxona” konsepsiyasi
doirasida tezkor diagnostika va monitoring uchun mo‘ljallangan bo‘lib, shovqinli muhitda ham
ishonchli natijalar beradi.
Havolalar
[1]. Smith, J., & Brown, T. (2021). Deep learning approaches for acoustic signal classification.
IEEE Transactions on Signal Processing, 69, 1234–1245.
[2]. Chen, L., et al. (2022). Automatic feature extraction in audio classification using
convolutional neural networks. IEEE Access, 10, 56789–56801.
[3]. Zhang, Y., & Wang, H. (2023). Time-frequency analysis of acoustic signals using deep
learning methods. Signal Processing, 198, 108567.
[4]. Goodfellow, I., Bengio, Y., & Courville, A. (2016). Deep Learning. MIT Press.
[5]. Oppenheim, A. V., & Schafer, R. W. (2010). Discrete-Time Signal Processing. Prentice Hall.
Deng, L., & Yu, D. (2014). Deep learning: Methods and applications. Foundations and
Trends in Signal Processing, 7(3–4), 197–387.
[6]. Krizhevsky, A., et al. (2012). ImageNet classification with deep convolutional neural
582
networks. NIPS.
[7]. Hershey, S., et al. (2017). CNN architectures for large-scale audio classification. ICASSP.
Wang, S., et al. (2020). Noise-robust acoustic signal processing using deep neural networks.
[8]. IEEE Transactions on Audio, Speech, and Language Processing, 28, 1520–1532. ISO 80601-
2-61:2017. Medical electrical equipment – Pulse oximeters. ISO. Bishop, C. M. (2006).
Pattern Recognition and Machine Learning. Springer.
[9]. Lecun, Y., Bengio, Y., & Hinton, G. (2015). Deep learning. Nature, 521(7553), 436–444.
