Detection of bone tumours in radiographic images using neural networks

被引:17
作者
Egmont-Petersen, M
Pelikan, E
机构
[1] Leiden Univ, Ctr Med, Div Image Proc, Dept Radiol, NL-2300 RC Leiden, Netherlands
[2] Philips Med Syst, Sci Tech Dept, Hamburg, Germany
关键词
bone tumours; feature selection; feed-forward neural network; quality assessment; self-organising feature map; texture;
D O I
10.1007/s100440050026
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
We develop an approach for segmenting radiographic images of focal bone lesions possibly caused by bone tumour. A neural network is used to classify individual pixels by a convolution operation based on a feature vector. We design eight features chat characterise the local texture in the neighbourhood of a pixel. Four of the features are based on co occurrence matrices computed from the neighbourhood. The true class label of the pixels in the radiographs are obtained from annotations made by an experienced radiologist. Neural networks and self-organising feature maps are trained to perform the segmentation cask. The experiments confirm che feasibility of using a feature-based neural network for finding pathologic bone changes in radiographic images. An analysis of the eight features indicates that the presence of edges and transitions, the complexity of the texture, as well as the amount of high frequencies in che texture, are che main features discriminating (soft) tissue from pathologic bone, the two classes most likely to be confused.
引用
收藏
页码:172 / 183
页数:12
相关论文
共 40 条
[1]   Signature verification using multiple neural classifiers [J].
Bajaj, R ;
Chaudhury, S .
PATTERN RECOGNITION, 1997, 30 (01) :1-7
[2]  
BOHNDORF K, 1993, ZENTRALBLATT RADIOLO, V147, P987
[3]   NEAREST NEIGHBOR PATTERN CLASSIFICATION [J].
COVER, TM ;
HART, PE .
IEEE TRANSACTIONS ON INFORMATION THEORY, 1967, 13 (01) :21-+
[4]   On the quality of neural net classifiers [J].
Egmont-Petersen, M. ;
Talmon, J.L. ;
Brender, J. ;
McNair, P. .
Artificial Intelligence in Medicine, 1994, 6 (05) :359-381
[5]   Assessing the importance of features for multi-layer perceptrons [J].
Egmont-Petersen, M ;
Talmon, JL ;
Hasman, A ;
Ambergen, AW .
NEURAL NETWORKS, 1998, 11 (04) :623-635
[6]  
EGMONTPETERSEN M, 1996, NETWORKS MED
[7]   NEURAL-NETWORK DIAGNOSIS OF MALIGNANT-MELANOMA FROM COLOR IMAGES [J].
ERCAL, F ;
CHAWLA, A ;
STOECKER, WV ;
LEE, HC ;
MOSS, RH .
IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, 1994, 41 (09) :837-845
[8]  
Freyschmidt J, 1988, KNOCHENTUMOREN KLIN
[9]   ON THE APPROXIMATE REALIZATION OF CONTINUOUS-MAPPINGS BY NEURAL NETWORKS [J].
FUNAHASHI, K .
NEURAL NETWORKS, 1989, 2 (03) :183-192
[10]  
Hand D.J., 1981, DISCRIMINATION CLASS