A consistent, quantitative, observer-independent method of characterising mammographic parenchymal pattern is described. The method is based on the calculation of the 'fractal dimension' of digitised mammograms. The degree of correlation between the parenchymal pattern classifications by a fractal-based system and those of radiologists is assessed. For a set of 70 mammograms, average weighted proportion agreement among three radiologists in calling Wolfe grades was 85%, while agreement between the radiologists and our fractal classifier was 84%. The method developed may prove to be useful in establishing an index of risk for breast cancer and, ultimately, in determining intervals between examinations for individuals in a mammographic screening programme.

Identification of the factors responsible for high-risk histological changes may offer new insights into the etiology of breast cancer and potentially lead to the development of methods for its prevention.

Summary.-We have carried out a case-control study to examine the relationship between mammographic signs and breast cancer. The mammographic signs assessed were prominent ducts and dysplasia.The cases were a group of 183 women with histologically verified unilateral breast cancer. The controls were a group of women attending a screening centre. Cases and controls were individually age-matched. Mammograms from the non-cancerous breast of the cases were randomly assembled with those of the controls and classified by 3 radiologists without knowledge of which films were from cases and which from controls.Mammographic dysplasia was found to be strongly associated with breast cancer, particularly in women aged < 50. Prominent ducts were only weakly associated with breast cancer. Multivariate analysis showed that the association between dysplasia and breast cancer could not be explained on the basis of other risk factors for breast cancer, and that classification of dysplasia discriminated more strongly between cases and controls than did classification of Wolfe's mammographic patterns.These results show that mammograms contain information about risk of breast cancer. Mammographic dysplasia is strongly associated with breast cancer, is present in a substantial proportion of patients with the disease, and may offer opportunities for prevention.

There is now a large amount of evidence indicating that women with extensive areas of mammographic densities are 4-6 times more likely to develop breast cancer than those with little or no density in the mammogram. We have examined one potential biological explanation for this association by estimating the incidence of various histological types of benign breast disease in relation to mammographic density. We studied the large cohort of women taking part in the National Breast Screening Study (NBSS), a randomized trial of screening with mammography. Mammograms from subjects with biopsies (n = 423) and from a comparison group of subjects randomly selected from the NBSS (n = 465) were included. Histological slides from biopsied subjects (n = 353) were classified independently by the pathologists of the NBSS and by a review pathologist (H.M.J.). Mammographic density in more than 75% of the breast area was associated with an increased risk of incidence of hyperplasia without atypia, and of atypical hyperplasia and/or carcinoma in situ. The classifications of the review pathologist showed that, compared to women with no density, the relative risk of incident lesions for women with density in more than 75% of breast was 13.85 (95% CI 2.65-72.49) for hyperplasia, and 9.23 (95% CI 1.66-51.48) for atypical hyperplasia and/or carcinoma in situ. These findings suggest that the association between extensive mammographic density and breast cancer risk may, at least in part, be attributable to biological processes in the breast that give rise to these histological features that are known to be related to breast cancer risk.

No abstract