Klasik ve klasik olmayan polikistik over sendromunun ayrımında MRG radyomik özelliklerin makine öğrenimine dayalı analizi

Year 2024, Volume: 49 Issue: 1, 89 - 96, 29.03.2024

Günay Rona , Neriman Zengin Fıstıkçıoğlu , Tekin Ahmet Serel , Meral Arifoğlu , Hanife Gülden Düzkalır , Şehnaz Evrimler , Serhat Özçelik , Kadriye Aydın

https://doi.org/10.17826/cumj.1393084

Abstract

Amaç: Bu çalışmanın amacı, klasik ve klasik olmayan polikistik over sendromunu (PKOS) ayırmada T2 ağırlıklı Manyetik Rezonans görüntüleme (MRG) görüntüleri üzerinde radyomik analizin değerini araştırmaktır.
Gereç ve Yöntem: Çalışmaya 2014-2022 yılları arasında pelvik MRG çekilen 101 PKOS hastasına ait (ortalama yaş 23±4) 202 over dahil edildi. Hastaların 53'ü (%52,5) fenotip A, 12'si (%11,9) fenotip B, 25'i fenotip C (%25,1) ve 11'i (%10,9) fenotip D idi. Overlerin 130'u (%64,4) klasik PKOS, 72'si (%35,6) klasik olmayan PKOS idi. Overler 3D Slicer programı kullanılarak tüm aksiyel kesitlerde manuel olarak segmente edildi. Toplam 851 özellik çıkarıldı. Makine öğrenimi (ML) analizi için Python 2.3, Pycaret Library programı kullanıldı. Veri kümeleri rastgele eğitim (%70, 141) ve test (%30, 61) veri kümelerine bölündü. ML algoritmalarının performansları AUC, doğruluk, hatırlama, kesinlik ve F1 puanlarıyla karşılaştırıldı.
Bulgular: Eğitim setindeki doğruluk ve AUC değerleri sırasıyla %57-%73 ve 0,50-0,73 arasındaydı. En iyi iki makine öğrenimi algoritması Random Forest (rf) (AUC:0,73, doğruluk: %73) ve Gradient Boosting Classifier (gbc) (AUC:0,71, doğruluk: %70) idi. Bu iki modelden elde edilen harman modelinin AUC, doğruluk, hatırlama ve kesinlik değerleri ile F1 puanı sırasıyla 0,70, %73, %56, %66, %58 olarak bulunmuştur.
Sonuç: T2A MR'dan elde edilen radyomik özellikler klasik ve klasik olmayan PKOS ayrımında faydalıdır.

Keywords

Polikistik over sendromu, fenotipler, manyetik rezonans görüntüleme, makine öğrenimi, radyomik, doku analizi

Machine learning-based analysis of MRI radiomics in the discrimination of classical and non-classical polycystic over syndrome

Abstract

Purpose: The aim of this study is to investigate the value of radiomics analysis on T2-weighted Magnetic Resonance imaging (MRI) images in differentiating classical and non-classical polycystic ovary syndrome (PCOS).
Materials and Methods: A total of 202 ovaries from 101 PCOS patients (mean age of 23±4 years) who underwent pelvic MRI between 2014 and 2022, were included in the study. Of the patients, 53 (52.5%) were phenotype A, 12 (11.9%) were phenotype B, 25 were phenotype C (25.1%), and 11 were phenotype D (10.9%). 130 (64.4%) of the ovaries were classical PCOS, 72 (35.6%) were non-classical PCOS. The ovaries were manually segmented in all axial sections using the 3D Slicer program. A total of 851 features were extracted. Python 2.3, Pycaret library was used for machine learning (ML) analysis. Datasets were randomly divided into train (70 %, 141) and test (30 %, 61) datasets. The performances of ML algorithms were compared with AUC, accuracy, recall, precision and F1 scores.
Results: Accuracy and AUC values in the training set ranged from 57%-73% and 0.50-0.73, respectively. The two best ML algorithms were Random Forest (rf) (AUC:0.73, accuracy:73%) and Gradient Boosting Classifier (gbc) (AUC:0.71, accuracy:70%). AUC, accuracy, recall and precision values and F1 score of the blend model obtained from these two models were 0.70, 73 %, 56 %, 66%, 58%, respectively.
Conclusion: Radiomic features obtained from T2W MRI are successful in distinguishing between classical and non-classical PCOS.

Keywords

Polycystic Ovary Syndrome, Phenotypes, Magnetic Resonance Imaging, Machine Learning, Radiomics, Texture Analysis

References

• Dumesic DA, Oberfield SE, Stener-Victorin E, Marshall JC, Laven JS, Legro RS. Scientific statement on the diagnostic criteria, epidemiology, pathophysiology, and molecular genetics of polycystic ovary syndrome. Endocr Rev. 2015;36:487-25.
• Azziz R. Polycystic Ovary Syndrome. Obstet Gynecol. 2018;132:321-36.
• Lizneva D, Suturina L, Walker W, Brakta S, Gavrilova-Jordan L, Azziz R. Criteria, prevalence, and phenotypes of polycystic ovary syndrome. Fertil Steril. 2016;106:6-15.
• Rotterdam ESHRE/ASRM-Sponsored PCOS Consensus Workshop Group. Revised 2003 consensus on diagnostic criteria and long-term health risks related to polycystic ovary syndrome. Fertility and sterility. 2004;81:19-25.
• Goodman NF, Cobin RH, Futterweit W, Glueck JS, Legro RS, Carmina E; American Association of Clinical Endocrinologists (AACE); American College of Endocrinology (ACE); Androgen Excess and PCOS Society (AES). American Association of Clinical Endocrinologists, American College of Endocrinology, And Androgen Excess And Pcos Society Disease State clinical review: guide to the best practices in the evaluation and treatment of polycystic ovary syndrome--part 1. Endocr Pract. 2015;21:1291-300.
• Dewailly D, Lujan ME, Carmina E, Cedars MI, Laven J, Norman RJ et al. Definition and significance of polycystic ovarian morphology: a task force report from the Androgen Excess and Polycystic Ovary Syndrome Society. Hum Reprod Update. 2014;20:334-52.
• Balen AH, Laven JS, Tan SL, Dewailly D. Ultrasound assessment of the polycystic ovary: international consensus definitions. Hum Reprod Update. 2003;9:505-14.
• Rotterdam ESHRE/ASRM-Sponsored PCOS Consensus Workshop Group. Revised 2003 consensus on diagnostic criteria and long-term health risks related to polycystic ovary syndrome. Fertil Steril. 2004;81:19-25.
• Brown M, Park AS, Shayya RF, Wolfson T, Su HI, Chang RJ. Ovarian imaging by magnetic resonance in adolescent girls with polycystic ovary syndrome and age-matched controls. J Magn Reson Imaging. 2013;38:689-93.
• Kenigsberg LE, Agarwal C, Sin S, Shifteh K, Isasi CR, Crespi R, Ivanova J et al. Clinical utility of magnetic resonance imaging and ultrasonography for diagnosis of polycystic ovary syndrome in adolescent girls. Fertil Steril. 2015;104:1302-9.e94.
• Kayemba-Kay's S, Pambou A, Heron A, Benosman SM. Polycystic ovary syndrome: Pelvic MRI as alternative to pelvic ultrasound for the diagnosis in overweight and obese adolescent girls. Int J Pediatr Adolesc Med. 2017;4:147-152.
• Yoo RY, Sirlin CB, Gottschalk M, Chang RJ. Ovarian imaging by magnetic resonance in obese adolescent girls with polycystic ovary syndrome: a pilot study. Fertil Steril. 2005;84:985-95.
• Fondin M, Rachas A, Huynh V, Franchi-Abella S, Teglas JP, Duranteau L et al. Polycystic Ovary syndrome in adolescents: which MR imaging-based diagnostic criteria?. Radiology. 2017;285:961-70.
• Pereira-Eshraghi CF, Tao R, Chiuzan CC, Zhang Y, Shen W, Lerner JP et al. Ovarian follicle count by magnetic resonance imaging is greater in adolescents and young adults with polycystic ovary syndrome than in controls. F S Rep. 2022;3:102-9.
• Aiyappan SK, Karpagam B, Vadanika V, Chidambaram PK, Vinayagam S, Saravanan KC. Age-related normogram for ovarian antral follicle count in women with polycystic ovary syndrome and comparison with age matched controls using magnetic resonance imaging. J Clin Diagn Res. 2016;10:11-3.
• Lambin P, Leijenaar RTH, Deist TM, Peerlings J, de Jong EEC, van Timmeren J et al. Radiomics: the bridge between medical imaging and personalized medicine. Nat Rev Clin Oncol. 2017;14:749-62.
• Rogers W, Thulasi Seetha S, Refaee TAG, Lieverse RIY, Granzier RWY, Ibrahim A, Keek SA, Sanduleanu S, Primakov SP, Beuque MPL, Marcus D, van der Wiel AMA, Zerka F, Oberije CJG, van Timmeren JE, Woodruff HC, Lambin P. Radiomics: from qualitative to quantitative imaging. Br J Radiol. 2020;93:2019-948.
• Yildiz, B.O., Bozdag, G., Harmanci, A. Otegen U, Boynukalin K, Vural Z et al. Increased circulating soluble P-selectin in polycystic ovary syndrome. Fertility and Sterility. 201;93:2311–15.
• Yildiz, B.O., Bolour, S., Woods, K. et al.Visually scoring hirsutism. Hum Reprod Update. 2010;16:51–64.
• Mumusoglu, S., Yildiz, B. O. Polycystic ovary syndrome phenotypes and prevalence: differential impact of diagnostic criteria and clinical versus unselected population. Current Opinion in Endocrine and Metabolic Research. 2020;12:66-71.
• Sachdeva G, Gainder S, Suri V, Sachdeva N, Chopra S. Comparison of the different PCOS phenotypes based on clinical metabolic, and hormonal profile, and their response to clomiphene. Indian J Endocrinol Metab. 2019;23:326-31.
• Razek AAKA, Elatta HA. Differentiation Between Phenotypes of Polycystic Ovarian Syndrome With Sonography. Journal of Diagnostic Medical Sonography. 2021;37:337-44.
• Lizneva D, Kirubakaran R, Mykhalchenko K, Suturina L, Chernukha G, Diamond MP et al. Phenotypes and body mass in women with polycystic ovary syndrome identified in referral versus unselected populations: systematic review and meta-analysis. Fertil Steril. 2016;106:1510-20.
• Carmina E, Campagna AM, Lobo RA. A 20-year follow-up of young women with polycystic ovary syndrome. Obstet Gynecol. 2012;119:263-9.
• Ladrón de Guevara A, Fux-Otta C, Crisosto N, Szafryk de Mereshian P, Echiburú B, Iraci G et al. Metabolic profile of the different phenotypes of polycystic ovary syndrome in two Latin American populations. Fertil Steril. 2014;101:1732-9.
• Franks S, McCarthy MI, Hardy K. Development of polycystic ovary syndrome: involvement of genetic and environmental factors. Int J Androl. 2006;29:278-90