Quantitative measurement of texture orientation in biomedical images using an artificial neural network

✍ Scribed by Zhen Zhang; Randall E. Scarberry; Marwan A. Simaan

• Publisher: John Wiley and Sons
• Year: 1998
• Tongue: English
• Weight: 172 KB
• Volume: 9
• Category: Article
• ISSN: 0899-9457
• DOI: 10.1002/(sici)1098-1098(1998)9:5<351::aid-ima5>3.0.co;2-9

✦ Synopsis

Texture orientation is one of the most important attriof heart tissue obtained through optical sectioning. The orientabutes used in biomedical and clinical image interpretation. It provides tions of image texture in these images provide the critical conneccritical clues of continuity and connectivity useful in relating adjacent tion that links tissue regions across the image sequence and helps image areas. We report a novel approach in which image data are tracking tissue fiber orientations in the 3D image volume data. convolved with directional convolution masks and the results are used In this article, a texture image area is considered to be oriented as input to an artificial neural network for classification of image areas when there exists a dominant direction in which image gray-level into a number of discrete texture orientation classes. ᭧ 1998 John distribution exhibits considerably less variability than in any other