REVISITING COMPARABILITY OF THE SIZES OF CHANGES CORRESPONDING TO BREAST CANCER IN FILM AND DIGITAL MAMMOGRAPHY

Abstract

In recent years, the traditional film technology of mammography has been replaced by digital technology, but comparative studies aimed at comparability of the size of changes when using these technologies are needed. The aim of the study was to assess the comparability of the size of changes corresponding to breast cancer in digital and film mammography, as well as the clinical significance of the dependencies identified. Material and methods. The study material was 27 sets of digital and film mammograms of patients with breast cancer, performed at intervals of no more than 2 months (digital mammograms – after film ones). The images were used to evaluate the same standardized neoplasm size. Results. The average size of changes corresponding to breast cancer on film mammograms was 16.96±8.44 mm, on digital ones – 15.63± 8.16 mm and 15.67±7.96 mm (p > 0.05). The range of variation in the difference in the neoplasm size according to film and digital mammography was 0-5 mm (0-30%). There were no cases of an increase in the neoplasm size on digital mammograms compared to film ones. In 37.04% (n = 10) of cases, the size of the changes in both types of mammography corresponded, in 29.63% (n = 8) understatement in digital mammography did not exceed 10%; in 25.93% (n = 7) of cases – 20%; in 7.40% (n = 2) – 30%. Underestimation of suspicious changes' sizes during digital mammography became clinically significant when their absolute size was less than 1 cm, when they became poorly visible against the background of fatty and fibrous-fatty parenchyma; as well as in situations when suspicious changes were indistinguishable from benign high-density zones. The use of a computer analysis system of mammograms in all cases ensured labeling changes corresponding to breast cancer. Conclusions. The sizes of changes corresponding to breast cancer (different from microcalcinates clusters, the size differences of which were not demonstrated) in film and digital mammography may differ by 0-30%, which does not allow using these techniques as interchangeable ones. Taking into account the possibility of underestimating the size of small changes, which makes them difficult to identify in the visual analysis of digital mammograms, it is advisable to use routinely a computer analysis system with subsequent targeted analysis of the labeled zones.