Man versus machine: Does automated computer density measurement add value?
Menée à partir de données mammographiques portant sur 1 609 patientes atteintes d'un cancer du sein, 351 patientes avec cancer mammaire invasif diagnostiqué entre deux examens de dépistage et sur 4 409 témoins, cette étude montre que la densité mammographique, qu'elle soit estimée par un radiologue ou mesurée par un système automatisé, peut prédire le risque de cancer de l'intervalle ou de cancer diagnostiqué dans le cadre d'un programme de dépistage
Among the many known risk factors for breast cancer (1), only breast density is known to decrease the accuracy of screening mammography. Breast density may obscure tumors, hampering a radiologist's ability to detect them. The risk for missed breast cancer is greatest in women with extremely dense breasts, and the rate of “interval cancer” detected in the year after a screening examination is higher in this setting. Density also is associated with increased rates of false-positive readings, leading to anxiety and additional testing (2). Laws in more than half of the United States now require that women be informed of their breast density, and some states also mandate that supplemental imaging with magnetic resonance or ultrasonography be recommended.
Annals of Internal Medicine , éditorial, 2017