Diagnosis Of Polycystic Ovary Syndrome With Xgboost Algorithm

Year 2024, Volume: 13 Issue: 3, 1234 - 1241, 26.09.2024

Ömer Çağrı Yavuz

https://doi.org/10.37989/gumussagbil.1366530

Abstract

Polycystic Ovary Syndrome (PCOS), a complex endocrine disorder, affects women of reproductive age. It involves a combination of symptoms including menstrual irregularities, hyperandrogenism and polycystic ovaries. Although the presence of an increased number of stroma and follicles on ultrasound indicates polycystic ovaries, this is not considered sufficient for the diagnosis of PCOS. Metabolic abnormalities, female pattern hair loss, sexual satisfaction and depression are associated with PCOS. Making sense of and analyzing these relationships is important for the diagnosis of PCOS. This study aims to diagnose PCOS with the XGBoost algorithm, which is frequently used in the literature in recent years and is reported to be faster and safer than other algorithms. In this direction, XGBoost algorithm was applied to the dataset taken from the Kaggle database and consisting of 554 records in total. The dataset used in the study was obtained from 10 different hospitals in Kerala (India). In addition, in order to examine the effects of categorical data on algorithm performance, different data sets were created and their performances were evaluated. Finally, the performance was tested by balancing the data set in order to reveal the effect of the distribution in the data set on the performance. With a dataset of 554 records, an accuracy value of 0,87 was obtained. In line with the performance metrics obtained in the study, it can be said that the XGBoost algorithm will contribute to the solution of classification problems in the field of health.

Keywords

Machine learning, polycystic ovary syndrome, classification

Xgboost Algoritmasıyla Polikistik Over Sendromu Teşhisi

Abstract

Karmaşık bir endokrin bozukluk olan Polikistik Over Sendromu (PKOS), üreme çağındaki kadınları etkilemektedir. Adet düzensizlikleri, hiperandrojenizm ve polikistik överler gibi çeşitli semptomların kombinasyonunu barındırır. Ultrasonda artan sayıda stroma ve folikül varlığı polikistik yumurtalıkları ifade etse de bu durum PKOS tanısı için yeterli görülmemektedir. Metabolik anormallikler, kadın tipi saç dökülmesi, cinsel tatmin ve depresyon PKOS ile ilişkilendirilmektedir. Bu ilişkilerin anlamlandırılması ve analiz edilmesi PKOS teşhisi için önem arz etmektedir. Bu çalışma kapsamında son yıllarda literatürde sıklıkla kullanılan ve diğer algoritmalara göre daha hızlı ve güvenli olduğu belirtilen XGBoost algoritmasıyla PKOS’un teşhis edilmesi amaçlanmıştır. Bu doğrultuda Kaggle veri tabanından alınmış ve toplamda 554 kayıttan oluşan veri setine XGBoost algoritması uygulanmıştır. Çalışmada kullanılan veri seti Kerala (Hindistan)'da yer alan 10 farklı hastaneden elde edilmiştir. Ayrıca kategorik verilerin algoritma performansı üzerindeki etkilerinin incelenmesi amaçlanarak farklı veri setleri oluşturularak performansları değerlendirilmiştir. Son olarak veri setindeki dağılımın performans üzerindeki etkisinin ortaya konulması amaçlanarak veri seti dengeli hale getirilerek performans test edilmiştir. 554 kayıttan oluşan veri setiyle 0,87 doğruluk değeri elde edilmiştir. Çalışmada elde edilen performans metrikleri doğrultusunda sağlık alanında sınıflandırma problemlerinin çözümünde XGBoost algoritmasının katkı sağlayacağı söylenebilir.

Keywords

Makine öğrenmesi, polikistik over sendromu, sınıflandırma

References

• 1. Russell, S.J. ve Norvig P. (2016). “Artificial intelligence: a modern approach”. Malaysia: Pearson Education Limited.
• 2. Muggleton S. (2014). “Alan Turing and the development of Artificial Intelligence”. AI communications, 27 (1), 3-10, 10,3233/AIC-130579
• 3. Machinery, C. (1950). “Computing machinery and intelligence-AM Turing”. Mind, 59 (236), 433.
• 4. McCarthy J., Minsky M.L., Rochester N. ve Shannon C.E. (2006). “A proposal for the dartmouth summer research project on artificial intelligence”. AI magazine, 27 (4), 12-12. https://doi.org/10,1609/aimag.v27i4.1904
• 5. Lewis, T. (2014). “A Brief History of Artificial Intelligence”. Live Science. Erişim adresi: https://www.livescience.com/49007-history-of-artificial-intelligence.html (Erişim tarihi: 23.07.2023).
• 6. Öztürk K. ve Şahin M.E. (2018). “Yapay sinir ağları ve yapay zekâ’ya genel bir bakış”. Takvim-i Vekayi, 6 (2), 25-36.
• 7. Rajeswari P., Sathishkumar V.E., Anilkumar C., Thilakaveni P. ve Moorthy U. (2023). “Big Data Analytics and Implementation Challenges of Machine Learning in Big data”. Applied and Computational Engineering, p.233-238.
• 8. Pachamanova D., Tilson V. ve Dwyer-Matzky K. (2022). “Case article—Machine learning, ethics, and change management: A data-driven approach to improving hospital observation unit operations”. INFORMS Transactions on Education, 22 (3). p.178-187. https://doi.org/10,1287/ited.2021.0251ca
• 9. Dewailly D. (2016) “Diagnostic criteria for PCOS: is there a need for a rethink?”. Best Practice & Research Clinical Obstetrics & Gynaecology, 37, 5-11. https://doi.org/10,1016/j.bpobgyn.2016.03.009
• 10. Michelmore K.F., Balen A.H., Dunger D.B. ve Vessey M.P. (1999). “Polycystic ovaries and associated clinical and biochemical features in young women”. Clinical endocrinology, 51 (6), 779-786. https://doi.org/10,1046/j.1365-2265.1999.00886.x
• 11. Ethirajulu A., Alkasabera A., Onyali C.B., Anim-Koranteng C., Shah H.E., Bhawnani N. ve Mostafa J.A. (2021). “Insulin resistance, hyperandrogenism, and its associated symptoms are the precipitating factors for depression in women with polycystic ovarian syndrome”. Cureus, 13 (9). 10,7759/cureus.18013
• 12. Aydos, A., Öztemur, Y. ve Dedeoğlu, B. G. (2016). “Polikistik over sendromu ve moleküler yaklaşımlar”. Türk Hijyen ve Deneysel Biyoloji Dergisi, 73(1), 81-88.
• 13. Jedel E., Gustafson D., Waern M., Sverrisdottir Y.B., Landen M., Janson P.O., Labrie F., Ohlsson C. ve Stener-Victorin E. (2011). “Sex steroids, insulin sensitivity and sympathetic nerve activity in relation to affective symptoms in women with polycystic ovary syndrome”. Psychoneuroendocrinology, 36 (10):1470–9. https://doi.org/10,1016/j.psyneuen.2011.04.001
• 14. Standeven L.R., Olson E., Leistikow N., Payne J.L., Osborne L.M. ve Hantsoo, L. (2021). “Polycystic ovary syndrome, affective symptoms, and neuroactive steroids: a focus on allopregnanolone”. Current psychiatry reports, 23 (6), 36. https://doi.org/10,1007/s11920-021-01244-w
• 15. Kuntal C., Vyas J., Chaudhary A., Hemani S. ve Rajoria L. (2021). “A study of metabolic syndrome in women with polycystic ovary syndrome at tertiary care center”. International Journal of Reproduction, Contraception, Obstetrics and Gynecology, 10 (6), 2427-2432.
• 16. Jiang V.S., Hawkins S.D. ve McMichael A. (2022). “Female pattern hair loss and polycystic ovarian syndrome: more than just hirsutism”. Current Opinion in Endocrinology & Diabetes and Obesity, 29 (6), 535-540, https://doi.org/10,1097/MED.0000000000000777
• 17. Kałużna M., Nomejko A., Słowińska A., Wachowiak-Ochmańska K., Pikosz K., Ziemnicka K. ve Ruchała M. (2021). “Lower sexual satisfaction in women with polycystic ovary syndrome and metabolic syndrome”. Endocrine Connections, 10 (9), 1035-1044. https://doi.org/10,1530/EC-21-0257 18. Sharma A. ve Verbeke W.J. (2020). “Improving diagnosis of depression with XGBOOST machine learning model and a large biomarkers Dutch dataset (n= 11,081)”. Frontiers in big Data, 3, 15. https://doi.org/10,3389/fdata.2020,00015
• 19. Chen T. ve Guestrin C. (2016). “Xgboost: A scalable tree boosting system”. In Proceedings of the 22nd acm sigkdd international conference on knowledge discovery and data mining, 785-794. https://doi.org/10,1145/2939672.2939785
• 20. Dhaliwal S.S., Nahid A.A. ve Abbas R. (2018). “Effective intrusion detection system using XGBoost”. Information, 9 (7), 149. https://doi.org/10,3390/info9070149.
• 21. Ning Y.L., Sun C., Xu X.H., Li L., Ke Y.J., Mai Y., ... Chen W.T. (2023). “Tendency of dynamic vasoactive and inotropic medications data as a robust predictor of mortality in patients with septic shock: An analysis of the MIMIC-IV database”. Frontiers in Cardiovascular Medicine. 10, 1126888. https://doi.org/10,3389/fcvm.2023.1126888
• 22. Chen S., Zhou W., Tu J., Li J., Wang B., Mo X., Tian G., Lv K. ve Huang Z. (2012). “A novel XGBoost method to infer the primary lesion of 20 solid tumor types from Gene expression data”. Frontiers in genetics, 12, 632761. https://doi.org/10,3389/fgene.2021.632761
• 23. Ogunleye A. ve Wang Q.G. (2019). “XGBoost model for chronic kidney disease diagnosis”. IEEE/ACM transactions on computational biology and bioinformatics, 17 (6), 2131-2140, 10,1109/TCBB.2019.2911071
• 24. Sharma N. (2018). “XGBoost. The extreme gradient boosting for mining applications”. Munich: GRIN Verlag.
• 25. Ramraj S., Uzir N., Sunil R. ve Banerjee S. (2016). “Experimenting XGBoost algorithm for prediction and classification of different datasets”. International Journal of Control Theory and Applications, 9(40), 651-662.
• 26. Ramaneswaran S., Srinivasan K., Vincent, P.D.R ve Chang C.Y. (2021). “Hybrid inception v3 XGBoost model for acute lymphoblastic leukemia classification”. Computational and Mathematical Methods in Medicine, 1-10, https://doi.org/10,1155/2021/2577375
• 27. Liew, X.Y., Hameed N. ve Clos J. (2021). “An investigation of XGBoost-based algorithm for breast cancer classification”. Machine Learning with Applications, 6, 100154. https://doi.org/10,1016/j.mlwa.2021.100154
• 28. Denny A., Raj A., Ashok A., Ram C.M. ve George R. (2019) “i-hope: Detection and prediction system for polycystic ovary syndrome (pcos) using machine learning techniques”. In TENCON 2019-2019 IEEE Region 10 Conference (TENCON), 673-678. IEEE.