یک روش مبتنی بر شبکه عصبی عمیق بهینه شده با الگوریتم هافمن و الگوریتم¬های فرا ابتکاری برای فشرده¬سازی و بازسازی تصویر پزشکی
محورهای موضوعی : مهندسی کامپیوتر و فناوری اطلاعات
محمد حسین خلیفه
1
(گروه مهندسی برق، واحد کازرون، دانشگاه آزاد اسلامی، کازرون، ایران)
مهدی تقی زاده
2
(هیات علمی)
محمدمهدی قنبریان
3
(کارشناس ارشد - دانشگاه آزاد اسلامی، واحد کازرون)
جاسم جمالی
4
(دانشگاه آزاد کازرون)
کلید واژه: فشرده¬سازی تصویر, بازسازی تصویر, شبکه عصبی عمیق, رمزنگاری هافمن, الگوریتم¬های فرا ابتکاری گرگ خاکستری ,
چکیده مقاله :
این تحقیق از دو رویکرد مختلف برای فشرده‌سازی عکس‌های پزشکی برای اهداف بلندمدت استفاده می‌کند. در روش اول، تصاویر با استفاده از رمز هافمن فشرده شده و سپس با استفاده از مدل‌سازی سلسله مراتبی بر اساس طبقه‌بندی طراحی شده توسط شبکه عصبی ساده‌سازی می‌شوند. در روش دوم از یک استراتژی پیش‌بینی مبتنی بر آموزش شبکه عصبی عمیق استفاده شده است. این روش از یک شبکه عصبی آموزش‌دیده برای استنتاج مکان‌های پیکسل‌های منفرد استفاده می‌کند و از این رو، مقدار داده‌های مورد نیاز برای توصیف یک تصویر را کاهش می‌دهد. رمزگذاری فشرده¬سازی هافمن روی داده¬های باقی¬مانده استفاده می¬شود. یک روش فیلتر فضایی پیشرفته برای رمزگشایی داده‌های تصویر استفاده می‌شود و سپس الگوریتم‌های فراابتکاری بهینه‌سازی اسب وحشی و بهینه‌سازی گرگ خاکستری برای تولید یک تصویر بازسازی‌شده استفاده می‌شوند. رویکردهای پیشنهادی امکان ساده‌سازی تصویر را فراهم می‌کنند که منجر به رمزگشایی سریع‌تر شده است. مدولاسیون شاخص تشابه ساختاری، زمان و نسبت سیگنال به نویز پیک به ترتیب به طور متوسط 2، 1/30 و 15/15 درصد نسبت به سایر روش¬ها بهبود یافته است. الگوریتم‌های پیشنهادی می‌توانند عکس‌های پزشکی را با کیفیت بسیار بالایی در مقایسه با روش‌های مبتنی بر یادگیری عمیق فعلی فشرده کنند.
This research makes use of two different approaches to compress medical images for long-term purposes. In the first method, images are compressed using the Huffman cipher and then simplified using a hierarchical modeling based on a neural network-designed categorization. A prediction strategy based on deep neural network training is employed in the second method. This technique uses a trained neural network to infer the locations of individual pixels, hence reducing the amount of data required to describe a picture. Huffman compression encryption is used on the leftover data. An enhanced spatial filtering technique is used to decode the picture data, and the wild horse optimization and gray wolf optimization meta-heuristic algorithms are then used to produce a rebuilt image. Without compromising compression efficiency, this allows for a more realistic application of the suggested solutions in non-deterministic contexts. The suggested approaches allow for picture simplification, which has resulted in faster decoding. Structural similarity index modulation, time and peak signal-to-noise ratio have been improved by an average of 2, 30.1 and 15.15%, respectively. The suggested algorithms were able to compress medical photos with very high quality level, as compared to the current deep learning-based methods.
[1] M.-A. Moreno-Ibarra, Y. Villuendas-Rey, M. D. Lytras, C. Yáñez-Márquez, and J.-C. Salgado-Ramírez, "Classification of diseases
using machine learning algorithms: A comparative study," Mathematics, vol. 9, p. 1817, 2021. [2] H. P. Chan, L. M. Hadjiiski, and R. K. Samala, "Computer‐aided diagnosis in the era of deep learning," Medical physics, vol. 47, pp. e218-e227, 2020
[3] W. Mbarki, M. Bouchouicha, S. Frizzi, F. Tshibasu, L. B. Farhat, and M. Sayadi, "Lumbar spine discs classification based on deep
convolutional neural networks using axial view MRI," Interdisciplinary Neurosurgery, vol. 22, p. 100837, 2020.
[4] J. Martinez-Mas, A. Bueno-Crespo, R. Martinez-Espana, M. Remezal-Solano, A. Ortiz-Gonzalez, S. Ortiz-Reina, et al., "Classifying Papanicolaou cervical smears through a cell merger approach by deep learning technique," Expert Systems with Applications, vol. 160, p. 113707, 2020.
[5] H. Zhou, K. Wang, and J. Tian, "Online transfer learning for differential diagnosis of benign and malignant thyroid nodules with
ultrasound images," IEEE Transactions on Biomedical Engineering, vol. 67, pp. 2773-2780, 2020. [6] F. Liu, M. Hernandez-Cabronero, V. Sanchez, M. W. Marcellin, and A. Bilgin, "The current role of image compression standards in
medical imaging," Information, vol. 8, p. 131, 2017. [7] J. Huang, S. Zhang, and D. Metaxas, "Efficient MR image reconstruction for compressed MR imaging," Medical Image Analysis,
vol. 15, pp. 670-679, 2011. [8] J. C. Salgado-Ramírez, J. M. Vianney Kinani, E. A. Cendejas-Castro, A. J. Rosales-Silva, E. Ramos-Díaz, and J. L. Díaz-de-Léon-
Santiago, "New model of heteroasociative min memory robust to acquisition noise," Mathematics, vol. 10, p. 148, 2022. [9] A. Benou, R. Veksler, A. Friedman, and T. R. Raviv, "Ensemble of expert deep neural networks for spatio-temporal denoising of
contrast-enhanced MRI sequences," Medical image analysis, vol. 42, pp. 145-159, 2017. [10] K.-H. Thung, P.-T. Yap, E. Adeli, S.-W. Lee, D. Shen, and A. s. D. N. Initiative, "Conversion and time-to-conversion predictions of mild cognitive impairment using low-rank affinity pursuit denoising and matrix completion," Medical image analysis, vol. 45, pp. 68-
82, 2018. [11] E. Agustsson, M. Tschannen, F. Mentzer, R. Timofte, and L. V. Gool, "Generative adversarial networks for extreme learned image
compression," in Proceedings of the IEEE/CVF International Conference on Computer Vision, 2019, pp. 221-231. [12] M. Li, W. Zuo, S. Gu, D. Zhao, and D. Zhang, "Learning convolutional networks for content-weighted image compression," in
Proceedings of the IEEE conference on computer vision and pattern recognition, 2018, pp. 3214-3223. [13] N. Johnston, D. Vincent, D. Minnen, M. Covell, S. Singh, T. Chinen, et al., "Improved lossy image compression with priming and spatially adaptive bit rates for recurrent networks," in Proceedings of the IEEE conference on computer vision and pattern
recognition, 2018, pp. 4385-4393. [14]
S. Santurkar, D. Budden, and N. Shavit, "Generative compression," in 2018 Picture Coding Symposium (PCS), 2018, pp. 258-262. [15] E. Agustsson, M. Tschannen, F. Mentzer, R. Timofte, and L. V. Gool, "Generative adversarial networks for extreme learned image
compression," in Proceedings of the IEEE/CVF International Conference on Computer Vision, 2019, pp. 221-231. [16] S. M. Belgaonkar and V. Singh, "Image compression and reconstruction in compressive sensing paradigm," Global Transitions
Proceedings, vol. 3, pp. 220-224, 2022. [17] Y. Pourasad and F. Cavallaro, "A novel image processing approach to enhancement and compression of X-ray images,"
International Journal of Environmental Research and Public Health, vol. 18, p. 6724, 2021. [18] J. Luo, J. Wu, S. Zhao, L. Wang, and T. Xu, "Lossless compression for hyperspectral image using deep recurrent neural networks,"
International Journal of Machine Learning and Cybernetics, vol. 10, pp. 2619-2629, 2019. [19] H. Zhang, F. Cricri, H. R. Tavakoli, N. Zou, E. Aksu, and M. M. Hannuksela, "Lossless image compression using a multi-scale
progressive statistical model," in Proceedings of the Asian Conference on Computer Vision, 2020. [20] D. Mishra, S. K. Singh, and R. K. Singh, "Lossy medical image compression using residual learning-based dual autoencoder model," in 2020 IEEE 7th Uttar Pradesh section international conference on electrical, electronics and computer engineering (UPCON),
2020, pp. 1-5. [21] J. Azar, A. Makhoul, R. Couturier, and J. Demerjian, "Robust IoT time series classification with data compression and deep
learning," Neurocomputing, vol. 398, pp. 222-234, 2020. [22] Z. Cheng, H. Sun, M. Takeuchi, and J. Katto, "Learned lossless image compression with a hyperprior and discretized gaussian mixture likelihoods," in ICASSP 2020-2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2020, pp.
2158-2162. [23] Z. Ma, H. Zhu, Z. He, Y. Lu, and F. Song, "Deep Lossless Compression Algorithm Based on Arithmetic Coding for Power Data,"
Sensors, vol. 22, p. 5331, 2022. [24] H. Fu, F. Liang, B. Lei, N. Bian, Q. Zhang, M. Akbari, et al., "Improved hybrid layered image compression using deep learning and
traditional codecs," Signal Processing: Image Communication, vol. 82, p. 115774, 2020. [25] H. Rhee, Y. I. Jang, S. Kim, and N. I. Cho, "Lossless image compression by joint prediction of pixel and context using duplex
neural networks," IEEE Access, vol. 9, pp. 86632-86645, 2021. [26] T. Malach, S. Greenberg, and M. Haiut, "Hardware-based real-time deep neural network lossless weights compression," IEEE
Access, vol. 8, pp. 205051-205060, 2020. [27] O. H. Nagoor, J. Whittle, J. Deng, B. Mora, and M. W. Jones, "Lossless compression for volumetric medical images using deep
neural network with local sampling," in 2020 IEEE International Conference on Image Processing (ICIP), 2020, pp. 2815-2819. [28] G. Xin and P. Fan, "A lossless compression method for multi-component medical images based on big data mining," Scientific
Reports, vol. 11, p. 12372, 2021. [29] A. Nasif, Z. A. Othman, and N. S. Sani, "The deep learning solutions on lossless compression methods for alleviating data load
on IoT nodes in smart cities," Sensors, vol. 21, p. 4223, 2021. [30] O. H. Nagoor, J. Whittle, J. Deng, B. Mora, and M. W. Jones, "Sampling strategies for learning-based 3D medical image
compression," Machine Learning with Applications, vol. 8, p. 100273, 2022. [31] S. Mirjalili, S. M. Mirjalili, A. Lewis, and G. W. Optimizer, "Advances in engineering software," Renewable and Sustainable Energy
Reviews, vol. 69, pp. 46-61, 2014. [32] I. Naruei and F. Keynia, "Wild horse optimizer: A new meta-heuristic algorithm for solving engineering optimization problems,"
Engineering with computers, vol. 38, pp. 3025-3056, 2022. [33] D. A. Huffman, "A method for the construction of minimum-redundancy codes," Proceedings of the IRE, vol. 40, pp. 1098-1101,
1952. [34]
B. O’Hanen and M. Wisan, "JPEG Compression'," ed: Citeseer, 2005. [35] P. Kaur, "Compression using fractional Fourier transform," a thesis submitted in the partial fulfillment of requirement for the
award of the degree of master of engineering in electronics and communicationDeemed University, 2010. [36] A. Odat, M. Otair, and M. Al-Khalayleh, "Comparative study between LM-DH technique and Huffman coding," International
Journal of Applied Engineering Research, vol. 10, pp. 36004-36011, 2015. [37]
J. Schmidhuber, "Deep learning in neural networks: An overview," Neural networks, vol. 61, pp. 85-117, 2015. [38] J.-T. Chien, "Chapter 7 - Deep Neural Network," in Source Separation and Machine Learning, J.-T. Chien, Ed., ed: Academic Press,
2019, pp. 259-320. [39] C. Cardenas, A. Mohamed, G. Sharp, M. Gooding, H. Veeraraghavan, and J. Yang, "Data from AAPM RT-MAC grand challenge
2019," The Cancer Imaging Archive, 2019. [40] C. E. Cardenas, A. S. Mohamed, J. Yang, M. Gooding, H. Veeraraghavan, J. Kalpathy‐Cramer, et al., "Head and neck cancer patient images for determining auto‐segmentation accuracy in T2‐weighted magnetic resonance imaging through expert manual
segmentations," Medical physics, vol. 47, pp. 2317-2322, 2020. [41] S. G. Jafarzadeh, M. N. A. Rahman, and D. A. Wahab, "Optimization of supply chain management based on response surface
methodology: A case study of iran khodro," World Applied Sciences Journal, vol. 20, pp. 620-627, 2012. [42] A. Deshpande, V. V. Estrela, and N. Razmjooy, Computational Intelligence Methods for Super-Resolution in Image Processing
Applications: Springer, 2021. [43] A. G. M. Strollo, E. Napoli, D. De Caro, N. Petra, and G. Di Meo, "Comparison and extension of approximate 4-2 compressors for
low-power approximate multipliers," IEEE Transactions on Circuits and Systems I: Regular Papers, vol. 67, pp. 3021-3034, 2020. [44] H. E.-d. H. Ahmed, H. M. Kalash, and O. S. Farag Allah, "Encryption quality analysis of the RC5 block cipher algorithm for digital
images," Optical Engineering, vol. 45, pp. 107003-107003-7, 2006. [45] S. M. Belgaonkar and V. Singh, "Image compression and reconstruction in compressive sensing paradigm," Global Transitions
Proceedings, vol. 3, pp. 220-224, 2022. [46] W. Al-Haidri, I. Matveev, M. A. Al-Antari, and M. Zubkov, "A Deep Learning Framework for Cardiac MR Under-Sampled Image
Reconstruction with a Hybrid Spatial and k-Space Loss Function," Diagnostics, vol. 13, p. 1120, 2023.