CNN Based Determination of Polycystic Ovarian Syndrome using Automatic Follicle Detection Methods

Year 2024, Volume: 27 Issue: 4, 1589 - 1601, 25.09.2024

Perihan Gülşah Gülhan , Güzin Özmen , Hüsnü Alptekin

https://doi.org/10.2339/politeknik.1263520

Abstract

The aim of this study was to determine the best method for follicle detection using ovarian ultrasound images and to classify the ultrasound images as pcos or normal ovaries using the proposed CNN architecture. Two different methods for follicle detection have been proposed to evaluate pcos. For this purpose, the Median, the Mean, the Wiener, and the Gaussian filters were tested using standard and adaptive thresholds. Second, Gaussian filtering, Discrete Wavelet Transform, and k-means clustering algorithms were tested. The Canny operator separates follicles from the background in the segmentation phase. In this study, a CNN architecture that classifies limited ultrasound ovary images was developed, and its success in the best follicle detection method was presented. The highest follicle detection accuracy of 97.63% was achieved with adaptive thresholding using a Wiener filter. Besides, the ultrasound images of the ovaries were classified as "normal" or "polycystic ovary syndrome" using CNN architecture with classification accuracy of 65.81% for unsegmented ovarian images and 77.81% for segmented images. In addition to the proposed method, classification was performed using SqueezeNet-based transfer learning, which was successful in limited datasets, and 74.18% classification accuracy was achieved for the unsegmented images and 75.54 % for segmented images . The results show that the combination of the Wiener filter with adaptive thresholding was quite successful in follicle detection and that the CNN can better classify ovaries using preprocessed ultrasound images.

Keywords

Automatic Follicle Detection, Classification, Convolutional Neural Networks, Image Preprocessing, Ultrasound Image.

Otomatik Folikül Saptama Yöntemleri Kullanılarak ESA Tabanlı Polikistik Over Sendromu Tespiti

Abstract

Bu çalışmanın amacı, yumurtalık ultrason görüntülerini kullanarak folikül tespiti için en iyi yöntemi belirlemek ve önerilen CNN mimarisini kullanarak ultrason görüntülerini pkos veya normal yumurtalık olarak sınıflandırmaktır. Pkos'u değerlendirmek için folikül tespitinde iki farklı yöntem önerilmiştir. Bu amaçla Ortanca, Ortalama, Wiener ve Gauss filtresi; standart ve uyarlanabilir eşikle test edilmiştir. İkinci olarak, Gauss filtreleme, Ayrık Dalgacık Dönüşümü ve k-means kümeleme algoritması test edilmiştir. Segmentasyon aşamasında folikülleri arka plandan ayırmak için Canny operatörü kullanılmıştır. Bu çalışmada sınırlı ultrason over görüntüsünü sınıflandıran CNN mimarisi geliştirilmiş ve mimarinin optimum folikül tespit yöntemindeki başarısı sunulmuştur. Wiener Filtresi kullanılarak uyarlanabilir eşikleme ile %97.63' lük en yüksek folikül tespit doğruluğu elde edilmiştir. Ayrıca CNN mimarisi kullanılarak yumurtalıkların ultrason görüntüleri "normal" veya "polikistik over sendromu" olarak segmente edilmemiş over görüntüleri için %65,81 ve segmente edilmiş görüntüler için %77.81 sınıflandırma doğruluğu sınıflandırılmıştır. Önerilen yöntemin yanı sıra, sınırlı veri kümesinde oldukça başarılı olan SqueezeNet tabanlı transfer öğrenme kullanılarak sınıflandırma yapıldı ve segmente edilmemiş görüntüler için %74,18, segmente edilmiş görüntüler için %75.54 sınıflama doğrulupu elde edildi. Sonuçlar, Wiener filtresinin uyarlamalı eşikleme ile kombinasyonunun folikül tespitinde oldukça başarılı olduğunu ve CNN'nin önceden işlenmiş ultrason görüntülerini kullanarak yumurtalıkları daha iyi sınıflandırabildiğini göstermektedir.

Keywords

Otomatik Folikül Tespiti, Sınıflandırma, Konvolüsyonel Sinir Ağları, Görüntü Ön İşleme, Ultrason Görüntüsü.

References

• [1] P. S. Hiremath and J. R. Tegnoor, “Follicle Detection and Ovarian Classification in Digital Ultrasound Images of Ovaries,” in Advancements and Breakthroughs in Ultrasound Imaging, Chapter 7: 167-197, (2013).
• [2] A. Gougeon, “Dynamics of Human Follicular Growth: Morphologic, Dynamic, and Functional Aspects,” The ovary, 2: 25-43, (2004).
• [3] T. D. Pache, J. W. Wladimiroff, W. C. Hop, and B. C. Fauser, “How to discriminate between normal and polycystic ovaries: transvaginal US study.,” Radiology, 183.2:421–423, (1992).
• [4] E. S. Knochenhauer, T. J. Key, M. Kahsar-Miller, W. Waggoner, L. R. Boots, and R. Azziz, “Prevalence of the Polycystic Ovary Syndrome in Unselected Black and White Women of the Southeastern United States: A Prospective Study1,” J Clin Endocrinol Metab, 83(9):3078–3082, (1998).
• [5] T. R. ESHRE and A.-S. P. C. W. Group, “Revised 2003 consensus on diagnostic criteria and long-term health risks related to polycystic ovary syndrome,” Fertil Steril, 19(1):19–25, (2004).
• [6] F. J. M. Broekmans, E. Knauff, O. Valkenburg, J. Laven, M. Eijkemans, and B. Fauser, “PCOS according to the Rotterdam consensus criteria: Change in prevalence among WHO-II anovulation and association with metabolic factors”, BJOG: An International Journal of Obstetrics & Gynaecology, 113(10):1210-1217, (2006).
• [7] S. M. Pfeifer, “Polycystic ovary syndrome in adolescent girls,” Semin Pediatr Surg, 14(2):111–117, (2005).
• [8] D. E. Lane, “Polycystic Ovary Syndrome and Its Differential Diagnosis,” Obstet Gynecol Surv, 61.2:125-135, (2006).
• [9] A. Dunaif, M. J. Graf, J. P. Mandeli, V. Laumas, and A. Dobrjansky, “Characterization of groups of hyperandrogenic women with acanthosis nigricans, impaired glucose tolerance, and/or hyperinsulinemia.,” J Clin Endocrinol Metab, 65(3):499–507, (1987).
• [10] A. Gougeon and B. Lefèvre, “Evolution of the diameters of the largest healthy and atretic follicles during the human menstrual cycle,” Reproduction, 69(2):497–502, (1983).
• [11] C. Battaglia et al., “Color Doppler analysis in oligo-and amenorrheic women with polycystic ovary syndrome,” Gynecological Endocrinology, 11(2):105–110, (1997).
• [12] A. H. Balen, J. S. E. Laven, S. Tan, and D. Dewailly, “Ultrasound assessment of the polycystic ovary: international consensus definitions,” Hum Reprod Update, 9(6):505–514, (2003).
• [13] M. J. Lawrence, M. G. Eramian, R. A. Pierson, and E. Neufeld, “Computer Assisted Detection of Polycystic Ovary Morphology in Ultrasound Images,” in Fourth Canadian Conference on Computer and Robot Vision, CRV ’07: 105–112, (2007).
• [14] P. Hiremath and J. Tegnoor, “Automatic Detection of Follicles in Ultrasound Images of Ovaries using Edge Based Method,” IJCA, Special Issue on RTIPPR 2, 120-125, (2010).
• [15] B. Purnama, U. N. Wisesti, Adiwijaya, F. Nhita, A. Gayatri, and T. Mutiah, “A classification of polycystic Ovary Syndrome based on follicle detection of ultrasound images,” in 2015 3rd International Conference on Information and Communication Technology (ICoICT), 396–401, (2015).
• [16] U. N. Wisesty, “Implementasi Gabor Wavelet dan Support Vector Machine pada Deteksi Polycystic Ovary (PCO) Berdasarkan Citra Ultrasonografi,” Indonesia Journal on Computing (Indo-JC), 1(2):67–82, (2016).
• [17] B. Padmapriya and T. Kesavamurthy, “Detection of follicles in poly cystic ovarian syndrome in ultrasound images using morphological operations,” Journal of Medical Imaging and Health Informatics, 6(1):240–243, (2016).
• [18] C. Sonigo et al., “High-throughput ovarian follicle counting by an innovative deep learning approach,” Scientific Reports, 8(1):13499, (2018).
• [19] A. Nazarudin, N. Zulkarnain, A. Hussain, S. Mokri, and I. N. A. Mohd Nordin, “Review on automated follicle identification for polycystic ovarian syndrome,” Bulletin of Electrical Engineering and Informatics, 9(2):588-593, (2020).
• [20] T. Zeng and J. Liu, “Automatic detection of follicle ultrasound images based on improved Faster R-CNN,” J Phys Conf Ser, 1187(4):042112, (2019).
• [21] M. Jayanthi Rao and R. Kiran Kumar, “Follicle Detection in Digital Ultrasound Images Using BEMD and Adaptive Clustering Algorithms,” in Innovative Product Design and Intelligent Manufacturing Systems: Select Proceedings of ICIPDIMS 2019, 651–659, (2020).
• [22] Ö. İnik, A. Ceyhan, E. Balcıoğlu, and E. Ülker, “A new method for automatic counting of ovarian follicles on whole slide histological images based on convolutional neural network,” Comput Biol Med, 112:103350, (2019).
• [23] B. Rachana, T. Priyanka, K. N. Sahana, T. R. Supritha, B. D. Parameshachari, and R. Sunitha, “Detection of polycystic ovarian syndrome using follicle recognition technique,” Global Transitions Proceedings, 2(2):304–308, (2021).
• [24] Adiwijaya, U. NOVIA WISESTY, and W. Astuti, “Polycystic Ovary Ultrasound Images Dataset.” Telkom University Dataverse, (2021).
• [25] P. G. YILMAZ and G. ÖZMEN, “Follicle Detection for Polycystic Ovary Syndrome by using Image Processing Methods,” International Journal of Applied Mathematics Electronics and Computers, 8(4):203–208, (2020).
• [26] R. C. Gonzalez, “Digital image processing”, Pearson education india, (2009).
• [27] Y. Zhu and C. Huang, “An improved median filtering algorithm for image noise reduction,” Phys Procedia, 25:609–616, (2012).
• [28] J. W. Tukey, “Exploratory data analysis”, 2: 131-160, (1977).
• [29] W. K. Pratt, Digital image processing: PIKS Scientific inside, Hoboken, New Jersey: Wiley-interscience, Vol.4, (2007).
• [30] M. M. P. Petrou and C. Petrou, “Image processing: the Fundamentals”, John Wiley & Sons, (2010).
• [31] P. Soille, “Morphological image analysis: principles and applications”, Springer, 2(3): 170-171, (1999).
• [32] N. Efford, “Digital image processing: a practical introduction using java (with CD-ROM)”, Addison-Wesley Longman Publishing, (2000).
• [33] N. I. Chervyakov, P. A. Lyakhov, and N. N. Nagornov, “Quantization noise of multilevel discrete wavelet transform filters in image processing,” Optoelectronics, Instrumentation and Data Processing, 54:608–616, (2018).
• [34] V. Kiruthika and M. M. Ramya, “Automatic segmentation of ovarian follicle using K-means clustering,” in 2014 fifth international conference on signal and image processing, 137–141, (2014).
• [35] B. Brahmadesam Viswanathan Krıshna, “Image pattern recognition technique for the classification of multiple power quality disturbances,” Journal of Electrical and Computer Sciences, 21(3):656–678, (2013).
• [36] G. Gupta, “Algorithm for image processing using improved median filter and comparison of mean, median and improved median filter,” International Journal of Soft Computing and Engineering (IJSCE), 1(5):304–311, (2011).
• [37] S. S. Khan and A. Ahmad, “Cluster center initialization algorithm for K-means clustering,” Pattern Recognit Lett, 25(11):1293–1302, (2004).
• [38] T. Brosnan and D.-W. Sun, “Improving quality inspection of food products by computer vision––a review,” Journal of Food Engineering, 61(1):3–16, (2004).
• [39] J. Schmidhuber, “Deep learning in neural networks: An overview,” Neural networks, 61:85–117, (2015).
• [40] W. You, C. Shen, D. Wang, L. Chen, X. Jiang, and Z. Zhu, “An intelligent deep feature learning method with improved activation functions for machine fault diagnosis,” IEEE Access, 8:1975–1985, (2019).
• [41] S. Minaee, R. Kafieh, M. Sonka, S. Yazdani, and G. J. Soufi, “Deep-COVID: Predicting COVID-19 from chest X-ray images using deep transfer learning,” Medical Image Analysis, 64:101794, (2020).
• [42] A. Hore and D. Ziou, “Image quality metrics: PSNR vs. SSIM,” in 2010 20th international conference on pattern recognition, IEEE, 2366–2369, (2010).
• [43] I. F. Thufailah and U. N. Wisesty, “An implementation of Elman neural network for polycystic ovary classification based on ultrasound images,” in Journal of Physics: Conference Series, 971(1):120168, (2018).