Image processing techniques for bone image analysis

Features of human bones are useful to establish correlation between bone structure and age, and information about age-related bone diseases. This paper presents a new approach to quantitative analysis of cross sections of human bones using image processing techniques. The system uses the adaptive neighborhood algorithm, clustering, local covariance measures, and the fuzzy region growing algorithm. The system extracts various bone features with consistency and provides more reliable statistics. As a result, the authors are able to correlate bone features with age and possibly with age related bone diseases such as osteoporosis.