Investigation of xeroradiography as a method of examination of the breast for cancer was begun a t Hutzel Hospital, Detroit, in 1966. T h e procedure appeared to have great promise. T h e following advantages of xeroradiography over conventional film mammography were: 1. ease of production; 2. ease of interpretation, and 3. clear recording of all densities of the breast on one image. I t was believed the method also would prove to be more accurate in diagnosis. A comparative study supported by a USPHS grant was designed a n d carried out over a 4-year period. Patients having mammography a t Hutzel Hospital were doubly examined, although care was taken not to exceed that amount of radiation required for the Egan technique of mammograph,hy. With the aid of statisticians, test sets were administered to radiologists in a few large cities in the U.S. Xeroradiography demonstrated a clear superiority in the diagnosis of malignant disease (84.3% xeroradiography vs. 72.2% film, true positive).
In benign disease, the two modes were very similar (75.5% xeroradiography vs. 75.0% film, true negative). Xeroradiography was significantly superior in overall accuracy (79.9% xeroradiography vs. 73.6% film). All figures are statistically significant a t least a t the 5% level or better.
STl!DY DESIGNED TO COXlPARE THE DIAG-A nostic utility of xeroradiography of the breast with conventional film mammography was conducted with the following results: seroradiography resulted in a significantly greater overall accuracy. Correct diagnoses 'were 17% greater than with the film technique on malignancies. In Ilenignancies, the rates .weye approximately equal with only a slight adv;mtage of xeroradiography. With all cases combined, xeroradiography was 9% inore accurate than film. In addition, the time required for interpretation of the xeroradiographs was 307; less than that required for Jilm, and 27 of the 36 radiologists participating in one part of the study expressed a prefcrence for the newer modality.