This paper deals with the detection of NOF fracture using statistical curve-fitting and Artificial Neural Network. It is difficult to detect NOF in pelvic X-ray images due to the complexity in the visual appearance of the pelvic region. An automatic algorithm is proposed in which statistical curve-fitting parameters are used for tracing the Shenton line in the pelvic image. The Artificial Neural Network is trained with Goodness of Fit parameters calculated using the statistical curve-fitting method for detecting the fracture in the image. The combination of statistical curve-fitting and ANN gives an accuracy of 96%.
Volume 11 | 10-Special Issue
Pages: 925-930
DOI: 10.5373/JARDCS/V11SP10/20192888