2 174

Cited 2 times in

Assessing the deformity of cleft lip nose based on neural network

Authors
 Wang Xi  ;  Felipe Vista IV  ;  Deok Won Kim  ;  Kil To Chong 
Citation
 INTERNATIONAL JOURNAL OF PRECISION ENGINEERING AND MANUFACTURING, Vol.11(3) : 473-482, 2010 
Journal Title
 INTERNATIONAL JOURNAL OF PRECISION ENGINEERING AND MANUFACTURING 
ISSN
 2234-7593 
Issue Date
2010
Keywords
Cleft lip nose ; Kendall’s coefficient ; Linear regression ; Neural Network
Abstract
The physical appearance of the nostril is important in the objective assessment of a cleft-lip patient while an objective quantitative evaluation is necessary to improve the result of the surgical procedure. The use of Kendall’s coefficient of concordance (W) to identify consistency between several raters is proposed in this paper. Linear regression method was then compared with the Neural Network method to find out which is better in determining the consistency of data. The feature factors were extracted from a digital image of the nostril taking into consideration symmetry as the basis. Statistical and Neural Network methods were utilized to process and analyze the deformity assessment data. Two groups of raters were chosen to evaluate the deformity of the cleft lip/ cleft nose based on photos shown to them. The angles and distance were measured with respect to the symmetrical aspect and the elementary reference score and factors were obtained through statistical analysis. Linear regression equations describing the relationship between the selected factors and the elementary score were formulated in order to obtain a more reliable reference data. The target data was pre-processed to achieve a more consistent and stable performance. A Neural Network was used to predict the evaluation score and it performed better than the linear regression method under certain conditions. The proposed method can give an objective evaluation to help surgeons evaluate their performance after a surgical procedure and find out if there is a need for further procedures to be done with lesser computational requirement over other existing three-dimensional algorithms.
Full Text
http://link.springer.com/article/10.1007%2Fs12541-010-0056-6#
DOI
10.1007/s12541-010-0056-6
Appears in Collections:
1. College of Medicine (의과대학) > Dept. of Medical Engineering (의학공학교실) > 1. Journal Papers
Yonsei Authors
Kim, Deok Won(김덕원)
URI
https://ir.ymlib.yonsei.ac.kr/handle/22282913/101253
사서에게 알리기
  feedback

qrcode

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

Browse