New Age Borders Obtained from Spot Photoscreener by Using Multivariate Cluster Analysis


Creative Commons License

BULUT E., ÇELİK Y., DAYI Ö., Bulut H.

European Journal of Therapeutics, cilt.28, sa.3, ss.203-208, 2022 (Hakemli Dergi) identifier identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 28 Sayı: 3
  • Basım Tarihi: 2022
  • Doi Numarası: 10.5152/eurjther.2022.21041
  • Dergi Adı: European Journal of Therapeutics
  • Derginin Tarandığı İndeksler: TR DİZİN (ULAKBİM)
  • Sayfa Sayıları: ss.203-208
  • Anahtar Kelimeler: Age border, classification, cluster analysis, dendrogram, photoscreener
  • İstanbul Gelişim Üniversitesi Adresli: Evet

Özet

Objective: The aim of this study is to analyze all the variables obtained from photoscreener using hierarchical cluster analysis to create more homogeneous age groups for more reliable and consistent measurement by photoscreener. Methods: The variables obtained from photoscreener examination of consecutive children who attended the ophthalmology department were evaluated. Medical records of the children were evaluated to ontain data including refractive parameters, deviation angle, pupil diameter variables and the calculated spherical equivalent, the cylindrical power vector J0, J45 values. These variables were analyzed by the multivariate cluster analysis. Results: Based on a dendrogram, 4 main clusters of similar quality variables were created. The calculated spherical equivalent decreased gradually from cluster I to IV, from 0.745 D to –0.235 D. The average pupil size in the Ist cluster was 5.06 mm, while in the IVth cluster, it was 6.38 mm. The proposed new age borders are distinct and statistically significant (P < .001). The ultimate proposed new age borders were found as 1-20, 21-64, 65-101, and 102-120 months, respectively. Conclusions: We proposed new age borders for the evaluation of refraction and pupil size of children which create new groups with a statistically different and homogeneous distribution. The proposed new age borders in this research would provide more reliable and consistent measurement results for clinical diagnosis.