Optical diffraction as a tool for semiautomatic, quantitative analysis of tissue specimens.

Optical diffraction was tested on electron micrographs of normal and malformed myelin sheaths as a method for semiautomatic quantitative analysis of tissue specimens. Both normal and malformed myelin sheaths were chosen for the analysis because of their characteristic internal structure and its alteration as a result of malformation. Optical diffraction patterns were obtained by means of an optical diffractometer coupled with a digital detector. The spacing and arrangement of the components of various types of myelin sheath were automatically calculated and determined and the results were verified with discriminant analysis. Out of 27 parameters of the radial and out of 25 parameters of the angular distributions of diffracted light intensity, 6 and 11, respectively, were found to have good discriminative power and were used for classification of myelin sheaths. The accuracy of automatic classification was tested by comparison with myelin sheath types of known origin. The samples visually similar by their appearance, e.g. control and regenerating myelin sheaths, were automatically classified with accuracy of 69%, whereas others were classified appropriately with 88-100% accuracy. It is believed that this kind of analysis may successfully be applied for specimens of other tissues and/or organs.