Introduction

The remarkable patterns exhibited by cancellous bone, and the likelihood that it is related to mechanical stress and strain, have been known for a long time (e.g., Todd and Bowman, 1845; Wyman, 1867; Wolff, 1892). During most of that period, however, the nature of the cancellous patterns themselves remained based upon visual impressions of bone sections and later, bone radiographs (e.g., Thompson, 1917; Murray, 1936; Evans, 1957). It was not until after the middle of the present century that the idea of characterizing these complex bony patterns with Fourier analyses came about.

In 1968, J. C. Davis showed me how useful optical Fourier transforms could be in the analysis of thin sections of rocks (Davis, 1970). In so doing he put into my mind the idea that the method might be excellent for the study of sectional and radiographic information in bones. Using his specialized equipment, we produced the first transforms of cancellous architecture in sections of human vertebrae (Oxnard, 1970a; 1970b; 1972a; 1972b; 1973). The 1973 volume may have been the first time that a Fourier transform was figured on the cover of a book.

Later H. C. Pincus (e.g., Power and Pincus, 1974) demonstrated to me the Rank Image Analyser 2,000 used for similar purposes.