Objective: To assess the predictive value of risk factors and ophthalmic examination findings for the development of retinopathy of prematurity (ROP) using artificial intelligence (AI) models. Material and Methods: A total of 453 premature infants between 22-33 weeks of gestation screened for ROP were evaluated retrospectively. The infants' perinatal risk factors (multiple births, small for gestational age, neonatal sepsis, etc) and ophthalmic examination findings were recorded. Random Forest model were trained with 10-fold cross validation using these variables to predict ROP. Accuracy, specificity, receiver operating characteristic curve, and area under the curve metrics were used to evaluate algorithm performance. Results: The model trained on all variables achieved 85% accuracy and 90% specificity in predicting ROP. On the other hand, the model trained on gestational age (GA), birth weight (BW) and perinatal risk factors achieved higher accuracy (87%) and specificity (90%) in predicting ROP compared to the model trained on GA and BW alone (76% accuracy and 82% specificity). When each variable was evaluated individually, the most effective factors were found to be total days on oxygen, GA, multiple birth and BW, respectively. In addition, the model was able to detect infants with stage II ROP (90% accuracy, 96% specificity) and zone III ROP (93% accuracy, 99% specificity) with higher accuracy and specificity. Conclusion: In addition to prematurity, exposure to perinatal risk factors is important in the development of ROP, and the evaluation of the effect of these factors using AI may support ROP specialists in the clinical management of infants.
Keywords: Artificial intelligence; retinopathy of prematurity; preterm; machine learning
Amaç: Bu çalışmanın amacı, yapay zekâ (YZ) modelleri kullanılarak prematüre retinopatisi [retinopathy of prematurity (ROP)] gelişiminde risk faktörleri ve oftalmik muayene bulgularının öngörücü değerinin değerlendirilmesidir. Gereç ve Yöntemler: ROP açısından tarama yapılan, 22-33. gebelik haftaları arasında doğan 453 prematüre yenidoğanın muayene bulguları geriye dönük değerlendirildi. Yenidoğanlara ait perinatal risk faktörleri (çoklu doğum, gestasyonel yaşa göre küçük doğum, neonatal sepsis vb.) ve oftalmolojik muayene bulguları kaydedildi. Bu değişkenler kullanılarak 10 katlı çapraz doğrulama ile ROP'u tahmin etmek için Rastgele Orman modeli eğitildi. Algoritma performansını değerlendirmek için doğruluk, özgüllük, alıcı işletim karakteristiği eğrisi ve eğrinin altındaki alan ölçümleri kullanıldı. Bulgular: Tüm değişkenler kullanılarak eğitilen modelin, ROP'u tahmin etmede %85 doğruluk ve %90 özgüllüğe ulaştığı görüldü. Gebelik yaşı (GY), doğum ağırlığı (DA) ve perinatal risk faktörleri ile eğitilen modelin, yalnızca GY ve DA ile eğitilen modele kıyasla (%76 doğruluk ve %82 özgüllük) ROP'u tahmin etmede daha yüksek doğruluk (%87) ve özgüllük (%90) elde ettiği tespit edildi. Her değişken tek tek değerlendirildiğinde en etkili faktörlerin sırasıyla toplam oksijen günü, GY, çoklu doğum ve DA olduğu bulundu. Ayrıca, modelin evre II ROP'lu (%90 doğruluk, %96 özgüllük) ve bölge III ROP'lu (%93 doğruluk, %99 özgüllük) bebekleri daha yüksek doğruluk ve özgüllükle tespit edebildiği görüldü. Sonuç: Prematüritenin yanı sıra perinatal risk faktörlerine maruziyet ROP gelişiminde önemlidir ve bu faktörlerin etkisinin YZ kullanılarak değerlendirilmesi, ROP uzmanlarına yenidoğanların klinik takibinde destek sağlayabilir.
Anahtar Kelimeler: Yapay zekâ; prematüre retinopatisi; prematüre; makine öğrenimi
- Kim SJ, Port AD, Swan R, Campbell JP, Chan RVP, Chiang MF. Retinopathy of prematurity: a review of risk factors and their clinical significance. Surv Ophthalmol. 2018;63(5):618-37. [PubMed] [PMC]
- Gilbert C, Fielder A, Gordillo L, Quinn G, Semiglia R, Visintin P, et al; International NO-ROP Group. Characteristics of infants with severe retinopathy of prematurity in countries with low, moderate, and high levels of development: implications for screening programs. Pediatrics. 2005;115(5):e518-25. [PubMed]
- Mora JS, Waite C, Gilbert CE, Breidenstein B, Sloper JJ. A worldwide survey of retinopathy of prematurity screening. Br J Ophthalmol. 2018;102(1):9-13. [PubMed]
- Slidsborg C, Jensen A, Forman JL, Rasmussen S, Bangsgaard R, Fledelius HC, et al. Neonatal risk factors for treatment-demanding retinopathy of prematurity: a danish national study. Ophthalmology. 2016;123(4):796-803. [PubMed]
- Mikolajczyk RT, Zhang J, Betran AP, Souza JP, Mori R, Gülmezoglu AM, et al. A global reference for fetal-weight and birthweight percentiles. Lancet. 2011;377(9780):1855-61. [PubMed]
- Gilbert C. Retinopathy of prematurity: a global perspective of the epidemics, population of babies at risk and implications for control. Early Hum Dev. 2008;84(2):77-82. [PubMed]
- Early Treatment For Retinopathy Of Prematurity Cooperative Group. Revised indications for the treatment of retinopathy of prematurity: results of the early treatment for retinopathy of prematurity randomized trial. Arch Ophthalmol. 2003;121(12):1684-94. [Crossref] [PubMed]
- Bolón-Canedo V, Ataer-Cansizoglu E, Erdogmus D, Kalpathy-Cramer J, Fontenla-Romero O, Alonso-Betanzos A, et al. Dealing with inter-expert variability in retinopathy of prematurity: a machine learning approach. Comput Methods Programs Biomed. 2015;122(1):1-15. [Crossref] [PubMed] [PMC]
- Valikodath N, Cole E, Chiang MF, Campbell JP, Chan RVP. Imaging in retinopathy of prematurity. Asia Pac J Ophthalmol (Phila). 2019;8(2):178-86. [Crossref] [PubMed] [PMC]
- Wakabayashi T, Patel SN, Campbell JP, Chang EY, Nudleman ED, Yonekawa Y. Advances in retinopathy of prematurity imaging. Saudi J Ophthalmol. 2022;36(3):243-50. [PubMed] [PMC]
- Ting DSW, Peng L, Varadarajan AV, Keane PA, Burlina PM, Chiang MF, et al. Deep learning in ophthalmology: the technical and clinical considerations. Prog Retin Eye Res. 2019;72:100759. [PubMed]
- Grzybowski A, Brona P, Lim G, Ruamviboonsuk P, Tan GSW, Abramoff M, et al. Artificial intelligence for diabetic retinopathy screening: a review. Eye (Lond). 2020;34(3):451-60. Erratum in: Eye (Lond). 2020;34(3):604. [PubMed] [PMC]
- Devalla SK, Liang Z, Pham TH, Boote C, Strouthidis NG, Thiery AH, et al. Glaucoma management in the era of artificial intelligence. Br J Ophthalmol. 2020;104(3):301-11. [PubMed]
- Reid JE, Eaton E. Artificial intelligence for pediatric ophthalmology. Curr Opin Ophthalmol. 2019;30(5):337-46. [PubMed]
- Brown JM, Campbell JP, Beers A, Chang K, Ostmo S, Chan RVP, et al; Imaging and Informatics in Retinopathy of Prematurity (i-ROP) Research Consortium. Automated diagnosis of plus disease in retinopathy of prematurity using deep convolutional neural networks. JAMA Ophthalmol. 2018;136(7):803-10. [PubMed] [PMC]
- Wang J, Ju R, Chen Y, Zhang L, Hu J, Wu Y, et al. Automated retinopathy of prematurity screening using deep neural networks. EBioMedicine. 2018;35:361-8. [PubMed] [PMC]
- Chen S, Zhao X, Wu Z, Cao K, Zhang Y, Tan T, et al. Multi-risk factors joint prediction model for risk prediction of retinopathy of prematurity. EPMA J. 2024;15(2):261-74. [PubMed] [PMC]
- Tapak L, Farahani LN, Taleghani NT, Ebrahimiadib N, Pour EK, Farahani AD, et al. Risk factors for the time to development of retinopathy of prematurity in premature infants in Iran: a machine learning approach. BMC Ophthalmol. 2024;24(1):364. [Crossref] [PubMed] [PMC]
- Koç E, Baş AY, Özdek Ş, Ovalı F. Türkiye Prematüre Retinopatisi Rehberi 2021 Güncellemesi. Türk Neonatoloji Derneği ve Türk Oftalmoloji Derneği. [Link]
- Chiang MF, Quinn GE, Fielder AR, Ostmo SR, Paul Chan RV, Berrocal A, et al. International classification of retinopathy of prematurity, third edition. Ophthalmology. 2021;128(10):e51-e68. [Crossref] [PubMed] [PMC]
- Good WV; Early Treatment for Retinopathy of Prematurity Cooperative Group. Final results of the Early Treatment for Retinopathy of Prematurity (ETROP) randomized trial. Trans Am Ophthalmol Soc. 2004;102:233-48; discussion 248-50. [PubMed] [PMC]
- Mintz-Hittner HA, Kennedy KA, Chuang AZ; BEAT-ROP Cooperative Group. Efficacy of intravitreal bevacizumab for stage 3+ retinopathy of prematurity. N Engl J Med. 2011;364(7):603-15. [PubMed] [PMC]
- Chen JS, Anderson JE, Coyner AS, Ostmo S, Sonmez K, Erdogmus D, et al. Quantification of early neonatal oxygen exposure as a risk factor for retinopathy of prematurity requiring treatment. Ophthalmol Sci. 2021;1(4):100070. [PubMed] [PMC]
- Masino AJ, Harris MC, Forsyth D, Ostapenko S, Srinivasan L, Bonafide CP, et al. Machine learning models for early sepsis recognition in the neonatal intensive care unit using readily available electronic health record data. PLoS One. 2019;14(2):e0212665. [PubMed] [PMC]
- Leigh RM, Pham A, Rao SS, Vora FM, Hou G, Kent C, et al. Machine learning for prediction of bronchopulmonary dysplasia-free survival among very preterm infants. BMC Pediatr. 2022;22(1):542. [PubMed] [PMC]
- Na JY, Kim D, Kwon AM, Jeon JY, Kim H, Kim CR, et al. Artificial intelligence model comparison for risk factor analysis of patent ductus arteriosus in nationwide very low birth weight infants cohort. Sci Rep. 2021;11(1):22353. [Crossref] [PubMed] [PMC]
- Cho H, Lee EH, Lee KS, Heo JS. Machine learning-based risk factor analysis of necrotizing enterocolitis in very low birth weight infants. Sci Rep. 2022;12(1):21407. [Crossref] [PubMed] [PMC]
- Gutierrez G. Artificial Intelligence in the Intensive Care Unit. Crit Care. 2020;24(1):101. [Crossref] [PubMed] [PMC]
- Agrawal R, Kulkarni S, Walambe R, Kotecha K. Assistive framework for automatic detection of all the zones in retinopathy of prematurity using deep learning. J Digit Imaging. 2021;34(4):932-47. [Crossref] [PubMed] [PMC]
- Tong Y, Lu W, Deng QQ, Chen C, Shen Y. Automated identification of retinopathy of prematurity by image-based deep learning. Eye Vis (Lond). 2020; 7:40. [Crossref] [PubMed] [PMC]
- Li P, Liu J. Early diagnosis and quantitative analysis of stages in retinopathy of prematurity based on deep convolutional neural networks. Transl Vis Sci Technol. 2022;11(5):17. [Crossref] [PubMed] [PMC]
- Huang YP, Basanta H, Kang EY, Chen KJ, Hwang YS, Lai CC, et al. Automated detection of early-stage ROP using a deep convolutional neural network. Br J Ophthalmol. 2021;105(8):1099-103. [Crossref] [PubMed] [PMC]
.: Process List