Evaluation of Termination Proposal for Pregnant Women Exposed to Radiation with Comparison of Machine Learning Methods

Creative Commons License

Kirişci M., Alay M. T.

10th International Eurasian Conference on Mathematical Sciences and Applications (IECMSA-2021), Sakarya, Türkiye, 25 - 27 Ağustos 2021, cilt.4, sa.3, ss.294-297, (Tam Metin Bildiri)

  • Yayın Türü: Bildiri / Tam Metin Bildiri
  • Cilt numarası: 4
  • Basıldığı Şehir: Sakarya
  • Basıldığı Ülke: Türkiye
  • Sayfa Sayıları: ss.294-297
  • Açık Arşiv Koleksiyonu: AVESİS Açık Erişim Koleksiyonu
  • İstanbul Üniversitesi-Cerrahpaşa Adresli: Evet

Özet

Objectives: The objective of the study is to establish a model that can directly predict the termination recommendation of physicians to whom women refer to our center for teratological counseling making the most accurate model with the least possible variable. Study design: Machine learning methods were performed to comprehensively analyze predict the termination recommendation of physicians to whom women refer to our center for teratological counseling. Methods: We determined the week of exposure to radiation, exposure dose, day of admission to hospital, age, and referring institution as a dependent variables. We compared the model we created using these variables with Naïve Bayes, kNN, SVM, ANN, and Logistic regression methods. Results: We estimated the physicians’ termination recommendation with 80% accuracy in ANN method. Conclusion: By using this model, pregnant women exposed to radiation can be detected before termination and non-indicative general losses can be prevented with appropriate teratological counseling.