Bitte wählen Sie ihr Lieferland und ihre Kundengruppe
Image processing and recognition methods are useful in many fields. This paper presents the hybrid optical and digital method applied to recognition of pathological changes in bones involved by metabolic bone diseases. The trabecular bone structure, registered by X-Ray on the photographic film, is analyzed in the new type of computer controlled diffractometer. The set of image parameters, extracted from diffractogram is evaluated by statistical analysis. The synthetic image descriptors in discriminant space, constructed on the base of 3 training groups of images (control, osteoporosis and osteomalacia groups) by discriminant analysis, allow to recognize bone samples with degraded bone structure and to recognize the disease. About 89 % of images were classified correctly. This method after optimization process will be verified in medical investigations.